Book 7

I think that enough has been said on the subject of invention. For I have dealt not merely with the methods by which we may instruct the judge, but also with the means of appealing to his emotions. But just as it is not sufficient for those who are erecting a building merely to collect stone and timber and other building materials, but skilled masons are required to arrange and place them, so in speaking, however abundant the matter may be, it will merely form a confused heap unless arrangement be employed to reduce it to order and to give it connexion and firmness of structure.

Nor is it without good reason that arrangement is treated as the second of the five departments of oratory, since without it the first is useless. For the fact that all the limbs of a statue have been cast does not make it a statue: they must be put together; and if you were to interchange some one portion of our bodies or of those of other animals with another, although the body would be in possession of all the same members as before, you would none the less have produced a monster. Again even a slight dislocation will deprive a limb of its previous use and vigour, and disorder in the ranks will impede the movements of an army.

Nor can I regard as an error the assertion that order is essential to the existence of nature itself, for without order everything would go to wrack and ruin. Similarly if oratory lack this virtue, it cannot fail to be confused, but will be like a ship drifting without a helmsman, will lack cohesion, will fall into countless repetitions and omissions, and, like a traveller who has lost his way in unfamiliar country, will be guided solely by chance without fixed purpose or the least idea either of starting-point or goal.

The whole of this book, therefore, will be devoted to arrangement, an art the acquisition of which would never have been such a rarity, had it been possible to lay down general rules which would suit all subjects. Put since cases in the courts have always presented an infinite variety, and will continue to do so, and since through all the centuries there has never been found one single case which was exactly like any other, the pleader must rely upon his sagacity, keep his eyes open, exercise his powers of invention and judgment and look to himself for advice. On the other hand, I do not deny that there are some points which are capable of demonstration and which accordingly I shall be careful not to pass by.

Division, as I have already stated, means the division of a group of things into its component parts, partition is the separation of an individual whole into its elements, order the correct disposition of things in such a way that what follows coheres with what precedes, while arrangement is the distribution of things and parts to the places which it is expedient that they should occupy.

But we must remember that arrangement is generally dependent on expediency, and that the same question will not always be discussed first by both parties. An example of what I mean, to quote no others, is provided by Demosthenes and Aeschines, who adopt a different order in the trial of Ctesiphon, since the accuser begins by dealing with the legal question involved, in which he thought he had the advantage, whereas the advocate for the defence treats practically every other topic before coming to the question of law, with a view to preparing the judges for a consideration of the legal aspect of the case.

For it will often be expedient for the parties to place different points first; otherwise the pleading would always be determined by the good pleasure of the prosecution. Finally, in a case of mutual accusation, where both parties have to defend themselves before accusing their antagonist, the order of everything must necessarily be different. I shall therefore set forth the method adopted by myself, about which I have never made any mystery: it is the result in part of instruction received from others, in part of my own reasoning.

When engaged in forensic disputes I made it a point to make myself familiar with every circumstance connected with the case. (In the schools, of course, the facts of the case are definite and limited in number and are moreover set out before we begin to declaim: the Greeks call them themes, which Cicero translates by propositions.) When I had formed a general idea of these circumstances, I proceeded to consider them quite as much from my opponent's point of view as from my own.

The first point which I set myself to determine (it is easy enough to state, but is still all-important) was what each party desired to establish and then what means he was likely to adopt to that end. My method was as follows. I considered what the prosecutor would say first: his point must either be admitted or controversial: if admitted, no question could arise in this connexion.

I therefore passed to the answer of the defence and considered it from the same standpoint: even there the point was sometimes one that was admitted. It was not until the parties ceased to agree that any question arose. 'fake for example the following case. You killed a man. Yes, I killed him. Agreed, I pass to the defence, which has to produce the motive for the homicide. It is lawful, lie urges, to kill an adulterer with his paramour. Another admitted point, for there is no doubt about the law. We must look for a third point where the two parties are at variance. They were not adulterers, say the prosecution; They were, say the defence. Here then is the question at issue: there is a doubt as to the facts, and it is therefore a question of conjecure. Sometimes even the third point may be admitted;

it is granted that they were adulterers. But, says the accuser, you had no right to kill them, for you were an exile or had forfeited your civil rights. The question is now one of law. On the other hand, if when the prosecution says, You killed them, the defence at once replies, I did not, the issue is raised without more delay. If it requires some search to discover where the dispute really begins, we must consider what constitutes the first question. The charge may be simple, as for example Rabirius killed Saturninus, or complex like the following: The offence committed by Lucius Varenus falls under the law of assassination for he procured the murder of Gaius Varenus, the wounding of Gnaeus Varenus and also the murder of Salarius. In the latter case there will be a number of propositions, a statement which also applies to civil suits as well. But in a complex case there may be a number of questions and bases: for instance the accused may deny one fact, justify another and plead technical grounds to show that a third fact is not actionable. In such cases the pleader will have to consider what requires refutation and where that refutation should be placed.

As regards the prosecutor, I do not altogether disagree with Celsus, who, though no doubt in so doing he is following the practice of Cicero, insists with some vehemence on the view that the first place should be given to some strong argument, but that the strongest should be reserved to the end, while the weaker arguments should be placed in the middle, since the judge has to be moved at the beginning and forcibly impelled to a decision at the end. But with the defence it is different:

the strongest arguments as a rule require to be disposed of first, for fear that the judge through having his thoughts fixed on those arguments should regard the defence of other points with disfavour. Sometimes, however, this order is subject to alteration; for example if the minor arguments are obviously false and the refutation of the most serious argument a matter of some difficulty, we should attack it last of all, after discrediting the prosecution by demonstrating the falsity of the former, thereby disposing the judges to believe that all their arguments are equally unreliable. We shall, however, require to preface our remarks by explaining why we postpone dealing with the most serious charge, and by promising that we will deal with it at a later stage: otherwise the fact that we do not dispose of it at once may give the impression that we are afraid of it.

Charges brought against the past life of the accused should generally be dealt with first in order that the judge may be well disposed to listen to our defence on that point on which lie has to give his verdict. But Cicero in the pro Vareno postpones his treatment of such charges to the conclusion, being guided not by the general rule, but by the special circumstances of the case.

When the accusation is simple, we must consider whether to give a single answer to the charge or several. In the former case, we must decide whether the question is one of fact or of law: if it is one of fact, we must deny the fact or justify it: if, on the other hand, it is a question of law, we must decide on what special point the dispute arises and whether the question turns on the letter or the intention of the law.

We shall do this by considering what the law is which gives rise to the dispute, that is to say under what law the court has been constituted. In scholastic themes, for example, the laws are sometimes stated merely with a view to connecting the arguments of the cases. Take the following case: A father who recognises a son whom he has exposed in infancy, shall only take him back after paying for his keep. A disobedient son may be disinherited. A man who took back a son whom he had exposed orders him to marry a wealthy neighbour. The son desires to marry the daughter of the poor man who brought him up.

The law about children who have been exposed affords scope for emotional treatment, while the decision of the court turns on the law of disinheritance. On the other hand, a question may turn on more laws than one, as in cases of ἀντινομία or contradictory laws. It is by consideration of such points as these that we shall be able to determine the point of law out of which the dispute arises.

As an example of complex defence I may quote the pro Rabirio: If he had killed him, he would have been justified in so doing: but he did not kill him. But when we advance a number of points in answer to a single proposition, we must first of all consider everything that can be said on the subject, and then decide which out of these points it is expedient to select and where to put them forward. My views on this subject are not identical with those which I admitted a little while ago on the subject of propositions and on that of arguments in the section which I devoted to proofs, to the effect that we may sometimes begin with the strongest.

For when we are defending, there should always be an increase of force in the treatment of questions and we should proceed from the weaker to the stronger, whether the points we raise are of the same or of a different character.

Questions of law will often arise from one ground of dispute after another, whereas questions of fact are always concerned with one point;

but the order to be followed is the same in both cases. We must, however, deal first with points that differ in character. In such cases the weakest should always be handled first, for the reason that there are occasions when after discussing a question we make a concession or present of it to our opponents: for we cannot pass on to others without dropping those which come first.

This should be done in such a way as to give the impression not that we regard the points as desperate, but that we have deliberately dropped them because we can prove our case without them. Suppose that the agent for a certain person claims the interest on a loan as due under an inheritance. The question may here arise whether such a claim can be made by an agent. Assume that, after discussing the question, we drop it or that the argument is refuted. We then raise the question whether the person in whose name the action is brought has the right to employ an agent. Let us yield this point also. The case will still admit of our raising the question whether the person in whose name the suit is brought is heir to the person to whom the interest was due and again whether he is sole heir.

Grant these points also and we can still raise the question whether the sum is due at all? On the other hand, no one will be so insane as to drop what he considers his strongest point and pass to others of minor importance. The following case from a scholastic theme is of a similar character. You may not disinherit your adopted son. And if you may disinherit him quâ adopted son, you may not disinherit one who is so brave. And if you may disinherit one who is so brave, you may not disinherit him because he has not obeyed your every command; and if he was bound to obey you in all else, you may not disinherit him on the ground of his choice of a reward; and even if the choice of a reward may give just ground for disinheriting, that is not true of such a choice as he actually made.

Such is the nature of dissimilarity where points of law are concerned. Where, however, the question is one of fact, there may be several points all tending to the same result, of which some may be dropped as not essential to the main issue, as for instance if a man accused of theft should say to his accuser, Prove that you had the property, prove that you lost it, prove that it was stolen, prove that it was stolen by me. The first three can be dropped, but not the last. I used also to employ the following method.

I went back from the ultimate species (which generally contains the vital point of the case) to the first general question or descended from the genus to the ultimate species, applying this method even to deliberative themes.

For example, Numa is deliberating whether to accept the crown offered him by the Romans. First he considers the general question, Ought I to be a king? Then, Ought I to be king in a foreign state? Ought I to be king at Rome? Are the Romans likely to put up with such a king as myself? So too in controversial themes. Suppose a brave man to choose another man's wife as his reward. The ultimate species is found in the question whether lie is allowed to choose another man's wife. The general question is whether he should be given whatever he chooses. Next come questions such as whether he can choose his reward from the property of private individuals, whether he can choose a bride as his reward, and if so, whether he can choose one who is already married.

But in our search for such questions we follow an order quite different from that which we employ in actual speaking. For that which as a rule occurs to us first, is just that which ought to come last in our speech: as for instance the conclusion, You have no right to choose another man's wife. Consequently undue haste will spoil our division of the subject. We must not therefore be content with the thoughts that first offer themselves, but should press our inquiry further till we reach conclusions such as that he ought not even to choose a widow: a further advance is made when we reach the conclusion that be should choose nothing that is private property, or last of all we may go back to the question next in order to the general question, and conclude that he should choose nothing inequitable.

Consequently after surveying our opponent's proposition, an easy task, we should consider, if possible, what it is most natural to answer first. And, if we imagine the case as being actually pleaded and ourselves as under the necessity of making a reply, that answer will probably suggest itself. On the other hand, if this is impossible, we should put aside whatever first occurs to us and reason with ourselves as follows: What if this were not the case? We must then repeat the process a second and a third time and so on, until nothing is left for consideration. Thus we shall examine even minor points, by our treatment of which we may perhaps make the judge all the better disposed to us when we come to the main issue.

The rule that we should descend from the common to the particular is much the same, since what is common is usually general. For example, He killed a tyrant is common, while A tyrant was killed by his son, by a woman or by his wife are all particular.

I used also to note down separately whatever was admitted both by my opponent and myself, provided it suited my purpose, and not merely to press any admissions that he might make, but to multiply them by partition, as for example in the following controversial theme:— A general, who had stood against his father as a candidate and defeated him, was captured: the envoys who went to ransom him met his father returning from the enemy. He said to the envoys, 'You are too late.' They searched the father and found gold in his pockets. They pursued their journey and found the general crucified. He cried to them, ' Beware of the traitor.' The father is accused. What points are admitted by both parties? We were told that there had been treason and told it by the general. We try to find the traitor. You admit that you went to the enemy, that you did so by stealth, that you returned unscathed, that you brought back gold and had it concealed about your person.

For an act of the accused may sometimes be stated in such a way as to tell heavily against him, and if our statement makes a real impression on the mind of the judge, it may serve to close his ears to all that is urged by the defence. For as a general rule it is of advantage to the accuser to mass his facts together and to the defence to separate them. I used also, with reference to the whole material of the case, to do what I have already mentioned as being done with arguments, namely, after first setting forth all the facts without exception, I then disposed of all of them with the one exception of the fact which I wished to be believed. For example, in charges of collusion it may be argued as follows.

The means for securing the acquittal of an accused person are strictly limited. His innocence may be established, some superior authority may intervene, force or bribery may be employed, his guilt may be difficult to prove, or there may be collusion between the advocates. You admit that he was guilty; no superior authority intervened, no violence was used and you make no complaint that the jury was bribed, while there was no difficulty about proving his guilt. What conclusion is left to us save that there was collusion?

If I could not dispose of all the points against me, I disposed of the majority. It is acknowledged that a man was killed: but he was not killed in a solitary place, such as might lead me to suspect that he was the victim of robbers; he was not killed for the sake of plunder, for nothing was taken from him; he was not killed in the hope of inheriting his property, for he was poor: the motive must therefore have been hatred, since you are his enemy.

The task not merely of division, but of invention as well, is rendered materially easier by this method of examining all possible arguments and arriving at the best by a process of elimination. Milo is accused of killing Clodius. Either he did or did not do the deed. The best policy would be to deny the fact, but that is impossible. It is admitted then that he killed him. The act must then have been either right or wrong. We urge that it was right. If so, the act must have either been deliberate or under compulsion of necessity, for it is impossible to plead ignorance.

The intention is doubtful, but as it is generally supposed to have existed, some attempt must he made to defend it and to show that it was for the good of the state. On the other hand, if we plead necessity, we shall argue that the fight was accidental and unpremeditated. One of the two parties then must have lain in wait for the other. Which was it? Clodius without doubt. Do you see how inevitably we are led to the right method of defence by the logical necessity of the facts?

We may carry the process further: either he wished to kill Clodius, who lay in wait for him, or he did not. The safer course is to argue that he did not wish to kill him. It was then the slaves of Milo who did the deed without Milo's orders or knowledge. But this line of defence shows a lack of courage and lessens the weight of our argument that Clodius was rightly killed.

We shall therefore add the words, As every man would have wished his slaves to do under similar circumstances. This method is all the more useful from the fact that often we can find nothing to say that really pleases us and yet have got to say something. Let us therefore consider every possible point; for thus we shall discover what is the best line for us to pursue, or at any rate what is least bad. Sometimes, as I have already said in the appropriate context, we may make good use of the statement of our opponent, since occasionally it is equally to the purpose of both parties. I am aware that some authors have written thousands of lines to show how we may discover which party ought to speak first. But in the actual practice of the courts this is decided either by some brutally rigid formula, or by the character of the suit, or finally by lot.

In the schools, on the other hand, such an enquiry is mere waste of time, since the prosecution and the defence are indifferently permitted to state a case and refute it in the same declamation. But in the majority of controversial themes it is not even possible to discover who should speak first, as for instance in the following: A certain man had three sons, an orator, a philosopher and a physician. In his will he divided his property into four portions, three of which he distributed equally among his sons, while the fourth was to go to the son who rendered the greatest service to his country.

The sons dispute the point. It is uncertain who should speak first, but our course is clear enough. For we shall begin with the son whose role we assume. So much for the general rules by which we should be guided in making our division.

But how shall we discover those questions which present abnormal difficulty? Just as we discover reflexions, words, figures or the appropriate nuances of style, namely by native wit, by study and by practice. None the less it will be rare for anyone who is not a fool to fail to discover them, so long as he is content, as I have said, to accept nature for a guide.

Many, however, in their passionate desire to win a reputation for eloquence are content to produce showy passages which contribute nothing to the proof of their case, while others think that their enquiry need not proceed further than that which meets the eye. To make my meaning clearer, I will cite a solitary example from the controversial themes of the schools; it is neither novel nor complicated.

The man who refuses to appear in defence of his father when accused of treason shall be disinherited: the man who is condemned for treason shall be banished together with his advocate. A father accused of treason was defended by one son who was a fluent speaker, while another son, who was uneducated, refused to appear for him. The father was condemned and banished with his advocate. The uneducated son performed some heroic act and demanded as a reward the restoration of his father and brother. The father returned and died intestate. The uneducated son claims a portion of his estate, the orator claims the whole for himself.

In this case those paragons of eloquence, who laugh at us because we trouble our heads about cases that rarely occur, will always assume the popular rôle. They will defend the uneducated against the eloquent son, the brave against the coward, the son who secured the recall of his kin against the ungrateful son, the son who is content with a portion of the inheritance against the son who would refuse his brother a share in their patrimony.

All these points are actually to be found in the case and are of considerable importance, but they are not such as to render victory a certainty. In such a case they will, as far as possible, search for daring or obscure reflexions (for to-day obscurity is accounted a virtue), and they will think they have given the theme a brilliant treatment by ranting and raving over it. Those, on the other hand, whose ideals are higher, but who restrict themselves merely to the obvious, will note the following points, which are, however, purely superficial.

The uneducated son may be excused for not appearing at the trial on the ground that he could contribute nothing to his father's defence: but even the orator has no claim on the gratitude of the accused, since the latter was condemned: the man who secured the recall of his kin deserves to receive the inheritance, while the man who refuses to divide it with his brother, more especially with a brother who has deserved so well of him, is avaricious, unnatural and ungrateful: they will further note that the first and essential question is that which turns on the letter and intention of the law; unless this is first disposed of, all subsequent arguments must fall to the ground.

He, however, who follows the guidance of nature will assuredly reflect as follows: the first argument of the uneducated son will be, My father died intestate and left two sons, my brother and myself; I claim a share in his estate by the law of nations. Who is so ignorant or so lacking in education as not to make this his opening, even though he does not know what is meant by a proposition?

He will then proceed to extol, though with due moderation, the justice of this common law of nations. The next point for our consideration is what reply can be made to so equitable a demand? The answer is clear:— There is a law which disinherits the man who fails to appear in his father's defence when the latter is accused of treason, and you failed to appear. This statement will be followed by the necessary praise of the law and denunciation of the man who failed to appear.

So far we have been dealing entirely with admitted facts. Let us now return to the claimant. Unless he is hopelessly unintelligent, surely the following argument will suggest itself:— If the law bars the way, there is no ground for action and the trial becomes a farce. But it is beyond question that the law exists and that the uneducated son did commit the offence for which it enacts a punishment. What then shall we say? I had no education.

But if the law applies to all men, it will be of no avail to plead lack of education. We must therefore try to discover whether there be not some point on which the law can be invalidated. We turn for guidance to nature (a point on which I cannot insist too often); what does she suggest save that when the letter of the law is against us, we should discuss its intention? This introduces the general question whether we are to stand by the letter or the spirit. But if we argue this question on general grounds with reference to law in the abstract, we shall go on for ever; it is a question that has never been decided. We must therefore restrict our enquiry to the particular law on which our case turns and try to find some argument against adhesion to the strict letter.

Well, then, is everyone who fails to appear in defence of his father to be disinherited? Are there no exceptions to the rule? At this point the following arguments will spontaneously suggest themselves. Is an infant liable to the law? For we may imagine a case where the son is an infant and has failed to appear in his father's support. Again does the law apply to a man who was away from home or absent on military service or on an embassy? We have gained a considerable amount of ground; for we have established the fact that a man may fail to appear for his father and still inherit.

Our declaimer, who has thought out this line of argument, must now pass over like a Latin flute-player, as Cicero says, to the side of the eloquent son and reply, Granted, but you are not an infant, you were not away from home nor absent on military service. Is there any answer to this except the previous reply, I am an uneducated man?

But to this there is the obvious retort, Even if you could not actually plead, you might have supported him by your presence, which is no more than the simple truth. The uneducated son must therefore return to the intention of the legislator. He wished to punish unfilial conduct, but I am not unfilial.

To this the eloquent son will reply, The action whereby you deserved disinheritance was unfilial, although penitence or desire for display may have subsequently led you to choose this as your reward. Further, it was owing to you that our father was condemned, since by absenting yourself you appeared to imply that you thought him guilty. The uneducated son replies, Nay, you contributed to his condemnation, for you had given offence to many and made our family unpopular. These arguments are based on conjecture, as also will be the excuse put forward by the uneducated son to the effect that his father advised his absence, as he did not wish to emperil his whole family. All these arguments are involved in the preliminary question as to the letter and the intention of the law.

Let us pursue the matter further and see if we can discover any additional arguments. How is that to be done? I am deliberately imitating the actual train of thought of one who is engaged in such an enquiry with a view to showing how such enquiry should be conducted. I shall therefore put aside the more showy kind of composition, and concern myself solely with such as may be of real profit to the student. So far we have derived all our questions from the character of the claimant. But why should we not make some enquiries into the character of the father? Does not the law say that whoever fails to appear for his father is to be disinherited?

Why should we not try asking whether this means that he is to be disinherited, whatever the character of the father for whom he failed to appear? Such a course is often adopted in those controversial themes in which we demand that sons who fail to maintain their parents should be cast into prison: take for example the case of the mother who gave evidence against her son when accused of being an alien, or of the father who sold his son to a procurer. What, then, is there in the present case that we lay hold of as regards the character of the father?

He was condemned. But does the law apply only to those cases where the father is acquitted? At first sight the question is difficult. But let us not despair. It is probable that the intention of the legislator was that innocent parents should secure the support of their children. But the uneducated son will be ashamed to produce this argument, since he acknowledges that his father was innocent.

There is, however, another line of argument which may be drawn from the enactment that the person condemned for treason should be banished together with his advocate. It seems almost impossible that in one and the same case a son should incur a penalty, both if he appeared in his father's defence and if he did not appear. Further, exiles are outlaws. Therefore the letter of the law cannot conceivably apply to the advocate of the condemned man.

For how can an exile hold any property? The uneducated son raises a doubt as to the interpretation both of the letter and the spirit of the law. Tile eloquent son will cling to the strict letter of tile law, which makes no exception, and will argue that the reason for enacting a penalty against those who fail to appear for their fathers was to prevent their being deterred from the defence of their fathers by the risk of banishment, and he will assert that his brother failed to appear in defence of his innocent father. It may therefore be worth while pointing out that two general questions may arise out of one basis — for we may ask, Is everyone who fails to appear liable to disinheritance? or Is he bound to appear irrespective of the character of his father?

So far all our questions have been derived from two of the persons involved. With regard to the third, this can give rise to no question, as there is no dispute about his portion of the inheritance. Still the time is not yet come to relax our efforts: for so far all the arguments might have been used even if the father had not been recalled from exile. But we must not betake ourselves at once to the obvious point that he was recalled by the agency of the uneducated son. A little ingenuity will lead us to look further a field: for as species comes after genus, so genus precedes species. Let us therefore assume that the father was recalled by someone else. This will give rise to a question of the ratiocinative or syllogistic type, namely whether recall from exile cancels the sentence of the court and is tantamount to the trial never having taken place at all. The uneducated son will therefore attempt to argue that, being entitled to not more than one reward, there was no means by which he could have secured the recall of his kin save by the restoration of his father on the same terms as if he had never been accused, and that this fact carries with it the cancellation of the penalty incurred by his advocate, as though he had never defended his father at all.

Our next point will be that which first occurred to us, namely the plea that he was recalled by the agency of the uneducated son. At this point we are confronted by the question whether the son who secured his father's restoration is thereby to be regarded in the light of an advocate, since he secured for him precisely what his original advocate demanded for him, and it is not an unreasonable claim to ask that an action should be regarded as equivalent when it is really more than equivalent.

The remaining points turn on questions of equity, for we ask which of the two sons makes the juster claim. This question admits of still further division. The claim of the uneducated son would have been the juster even if both had claimed the whole property. How much more so when one claims only a half and the other the whole to the exclusion of his brother. And then, even after we have dealt with all these points, an appeal to the memory of his father will carry great weight with the judges, more especially as the dispute is about the father's estate. This will give rise to conjecture as to what the intentions of the father were at the time of his dying intestate. This conjecture, however, involves a question of quality, and is employed in the service of a different basis.

As a rule questions of equity are best introduced at the conclusion of a case, since there is nothing to which the judges give more ready hearing. Sometimes, however, the interests of the case demand a change in this order; for example if we regard our case as weak in point of law, it will be well to secure the good-will of the judge by dealing with the question of equity first.

This concludes my general rules on this subject. We will now proceed to consider the several parts of forensic cases, and although I cannot follow them to the ultimate species, that is to say, I cannot deal with individual suits and controversies, I shall be able to discuss them on general lines in such a way as to show what bases most of them involve. And since the first question naturally is whether an alleged fact has taken place, I will begin with this.

All conjecture is concerned either with facts or intention. Each of these may occur in one of three times, past, present or future. Questions concerning facts are either general or definite, that is to say, those which involve consideration of persons and those which do not.

Concerning intentions there can be no questions which do not involve some person and where the facts of the case are not admitted. Therefore when the question turns on some fact, the point on which doubt arises is either what has been done, or what is being done, or what is likely to be done. For example, in general questions we discuss whether the universe has been formed of a concourse of atoms, or is governed by providence, or is likely some day to come to an end. In definite questions, on the other hand, we discuss whether Roscius has murdered his father, whether Manlius is aiming at making himself king, or Quintus Caecilius will be justified in appearing as the accuser of Verres.

In the law courts past time is of most importance, since all accusations are concerned with what has actually been done, while what is being done or is likely to be done is inferred from the past. We also enquire into origins. For instance, we enquire whether a pestilence be due to the anger of heaven, the inclement weather, the pollution of the water-supply, or the noxious vapours emitted by the earth. Again, we seek for the motives of an act. For example, we enquire whether the fifty kings who sailed against Troy did so because they were bound by their oath, or were moved to do so by righteous indignation, or merely desired to gratify the sons of Atreus. There is no very great difference between these two classes of question.

As regards facts falling within the present, if they can be detected by the eye without any reference to their logical antecedents being required, there will be no need of conjecture: let us suppose, for instance, that the Lacedaemonians are enquiring whether the Athenians are erecting fortifications. But although conjecture may seem entirely foreign to this class of question, there are cases in which it it necessary, as in questions of personal identity, which may be illustrated by the action brought against the heirs of Urbinia, where the question was whether the man who claimed the property as being the son of the deceased, was Figulus or Sosipater.

In this case the actual person was before the eyes of the court, so that there could be no question whether he existed (as there is, for instance, when we ask whether there exists any land beyond the Ocean) nor what he was nor of what kind. The question was simply, who he was. But this kind of dispute also depends on past time. The problem is whether this man Clusinius Figulus was born of Urbinia. Such disputes have arisen even in our own day, indeed I myself have pleaded in such. On the other hand, conjecture as to intention is obviously concerned with all three times. We ask with what purpose Ligarius went to Africa, with what purpose Pyrrhus is asking for a treaty, and how Caesar will take it if Ptolemy kills Pompey. We may also employ conjecture to enquire into quality in questions dealing with size, species and number, such as whether the sun is greater than the earth, whether the moon is spherical, flat or conical, whether there is one universe or several, or, to go outside these physical speculations, whether the Trojan or the Peloponnesian war was the greatest, what was the nature of the shield of Achilles, or whether there was more than one Hercules. In forensic cases, however, which consist of accusation and defence, there is one kind of conjecture by which we enquire both about an act and about its author. This sometimes treats the two questions together, as, for example, when both the act and the identity of the author are denied, and sometimes separately, as when the first enquiry, whether the act was committed, is followed by a second, where, the act being admitted, the question is by whom it was committed.

The act itself again sometimes involves a single question, as, for example, whether a man is dead, and sometimes two, as, for instance, whether he died of poison or of some internal disease. Another form of conjecture is concerned with the act alone, it being admitted that if the act was really committed, there can be no doubt as to its author. A third form is concerned solely with the author, the act being admitted and the dispute turning on the question as to who committed it. This third form is complex.

For the accused either confines himself to denying that he did it or accuses another of having done it. Further, there is more than one way of transferring the charge to another. At times this results in mutual accusation, which the Greeks call ἀντικατηγορία, and some of our own authors concertative accusation. At times, on the other hand, the charge is transferred to some person who cannot be brought to trial, and may be either known or unknown: again, if the person is known, he may be someone outside the case or the victim himself, who may be alleged to have committed suicide.

In such cases we compare characters, motives and other circumstances in the same way as in eases of mutual accusation. Cicero, for instance, in the pro Vareno diverts the charge from the accused to the slaves of Ancharius and in the pro Scauro throws the suspicion of Bostar's murder upon his mother.

There is also a different form of comparison, which comes into play when both parties claim the credit of some act, and yet another kind, when the question is not as between two persons, but as between two acts; that is to say, the question is not which of the two committed an act, but which of two acts was committed. Finally, if the act and the identity of the author are both admitted, we may still raise the question of his intention. I shall now proceed to detail. As an example of joint denial covering both the act and the identity of the author we may take the following statements, I have not committed adultery, I have not sought to establish myself as tyrant. In cases of murder or poisoning the denial is often divided as follows:

The act was not committed, and, if it was committed, it was not by me. But if the defence say, Prove that the man was killed, the burden falls solely on the accuser, for the accused can say nothing more against the charge except perhaps in the way of casting certain suspicions, which he should throw out in the vaguest terms, since if you make one definite assertion, you will have to prove it or run the risk of losing your case. For when the question lies between our statement and that of our opponent, one or other will be regarded as true. Thus when the point on which we relied for our defence is overthrown, there is nothing left but the points that tell against us.

For example, when the question turns on symptoms, which may point either to poisoning or internal disease, there is no third course left open and consequently each party sticks to his statement. At times the question turns on the nature of the fact, whether, for instance, death was due to poisoning or internal disease, and arguments are introduced which are drawn from the circumstances alone without any reference to the person concerned.

For example, it makes a difference whether the deceased was cheerful or depressed, had been working or taking his ease, had been awake or sleeping previous to the festive gathering that was followed by his death. The age of the deceased is also an important factor, and it is desirable to know whether he died suddenly or after a long period of ill health. If the question turns only on his sudden death, both parties will have still freer scope for discussion.

At times the character of the accused may be adduced to prove the fact, and to make it likely that it was or was not a case of poisoning because the accused is or is not a likely person to have committed such an act. When, on the other hand, the enquiry concerns both the accused and the act, the natural order for the accuser to pursue is to commence by proving that the act has been committed and then to go on to show that it was committed by the accused. If, however, proofs of the authorship of the crime are more in number than the proofs of the commission, this order may be reversed.

On the other hand, the accused will always begin by denying the act, since if this can be successfully proved, there is no need to say anything more, while if it is not proved, there remain other means of defence. Similarly, when the dispute turns solely on the act and, the act being proved, there can be no doubt as to the author, arguments may be drawn in like manner both from the person and the facts, although with reference to the question of fact alone.

Take the following controversial theme as an example, for it is best to employ scholastic themes as illustrations since they are more familiar to the student. A man who had been disinherited by his father took to the study of medicine. His father fell sick and, his life being despaired of by the other doctors, the son was called in, and said he would cure him if he would take a draught prescribed by himself. The father after drinking part of the draught said that he had been poisoned: the son drank the remainder of the draught. The father died and the son is accused of parricide.

There is no doubt who administered the draught, and, if it was poison, there is no question as to the author: but the problem as to whether the draught was poison can only be decided by arguments drawn from the character of the accused. There remains a third type of conjectural case where the fact is admitted, and the only question is as to the author. It is unnecessary for me to quote examples, since such cases are of frequent occurrence. For example, it may be clear that a man has been killed or that sacrilege has been committed, but the person accused of the crime may deny his guilt. It is from such circumstances that cases of mutual accusation arise, where it is admitted that the crime has been committed, but each party charges the other with being the author.

With regard to this class of case Celsus points out that they cannot actually occur in the courts, a fact which I imagine is familiar to all: for the jury is empanelled to try one accused person only, and even though the defence and the prosecution may accuse each other of tile crime, the first case must be tried before the second.

Apollodorus again stated that mutual accusation involved two separate disputes, and this is of course in conformity with the practice of the courts, which insists on two separate trials. On the other hand, mutual accusation is possible in cases tried before the senate or the emperor, and even in the courts the fact of mutual accusation will involve no difference in the pleadings, since the same methods will be required whether the verdict is given on both charges simultaneously or only on one.

In such cases the defence must always come first for three reasons. In the first place, we naturally prefer to secure our own safety than to injure our opponent, while secondly, our accusation will carry greater weight if we have first proved our own innocence, and thirdly, we shall thus secure a double line of defence. For the man who says, I did not kill him, is then free to go on to say, You killed him, whereas it is superfluous for the man who says, You killed him, to go on to say, I did not kill him.

Further, such cases consist of comparison, which may be effected in different ways. For we may either compare our case in its entirety with that of our adversary, or we may compare individual arguments. The choice between these two methods can only be determined by the requirements of the case. For example, in the pro Vareno, Cicero, in dealing with the first charge, compares the individual arguments: for it would have been rash in connexion with the second charge to compare the position of a stranger with that of a mother. It is therefore best, if possible, to refute argument by individual argument: if, however, our individual arguments are weak, we shall try to secure success by comparison of case with case as a whole.

But whether the case is one of mutual accusation, or the accused throws the guilt upon his opponent without making any formal accusation (as Roscius did without indicting his accusers), or the responsibility for the deed be placed on the victims themselves, whom we allege to have perished by their own hand, the arguments for both sides of the case will be compared in exactly the same way as in cases of mutual accusation.

The class of case last mentioned by me is, however, not merely of frequent occurrence in the schools, but sometimes actually occurs in the courts. For example, the sole question in the case of Naevius of Arpinum was whether he threw his wife out of the window or she threw herself. My speech in this case is the only one of all my pleadings that I have so far published, and I admit that I was led to do so merely by a youthful desire for glory. For the other speeches which circulate as mine have little in them that actually fell from my lips, having been corrupted by the carelessness of the shorthand-writers who took them down with a view to making money out of them.

There is also another type of conjectural case which, though it involves two questions, is different from cases of mutual accusation; such cases are concerned with rewards and may be illustrated by the following controversial theme. A tyrant, suspecting that his physician had given him poison, tortured him and, since he persisted in denying that he had done so, sent for a second physician. The latter asserted that poison had been administered, but that he would provide an antidote; he gave him a draught: the tyrant drank it and died. Both physicians claim a reward for slaying the tyrant. Now just as in cases of mutual accusation where each party shifts the guilt to his opponent, so in this case we compare the characters, motives, means, opportunities, instruments and evidence of the persons who claim the reward.

There is yet another type of case which, though not one of mutual accusation, is treated in the same way: I mean a case in which we enquire, without accusing anyone, which of two acts has taken place. For both parties make and defend their own statement of the case. Thus in the suit concerning the estate of Urbinia the claimant says that Clusinius Figulus, the son of Urbinia, on the defeat of the army in which he was serving, fled and after various misfortunes, being even even kept in captivity by the king, at length returned to Italy and his own home in the Marrucine district, where he was recognised. To this Pollio replies that he had been a slave to two masters at Pisaurum, that he had practised medicine, and finally, after receiving his freedom, inserted himself into a gang of slaves who were for sale and was at his own request purchased by himself.

Does not the whole suit consist of comparison between the two cases and of two different and opposite sets of conjecture? But the method to be followed is identical whether the case be one of accusation and defence or of claim and denial of the claim. Conjecture is, in the first place, based on what is past, under which I include persons, causes and intent. For in dealing with a case we first ask what the accused intended to do, next what he was in a position to do, and lastly what he actually did. Consequently the first point on which we must fix our attention is the character of the accused.

It is the business of the accuser to make any charge that he may bring against the accused not merely discreditable, but as consistent as possible with the crime for which he is arraigned. For example, if he calls a man accused of murder a debauchee or an adulterer, the discredit attaching to such charges will no doubt tell against the accused, but will, on the other hand, do less to prove the case than if he shows him to be bold, insolent, cruel or reckless.

On the other hand, counsel for the defence must, as far as possible, aim at denying, excusing or extenuating such charges, or, if that be impossible, show that they are not relevant to the case. For there are many charges which not only have no mutual resemblance, but may even at times contradict each other, as for instance if a man accused of theft is called prodigal or careless. For it is not likely that one and the same man should at once despise money and covet it.

If such means of defence are not available, we must take refuge with the plea that the charges made are not relevant to the case, that because a man has committed certain sins, it does not follow that he has committed all, and that the accusers ventured to make such false charges merely because they hoped by injuring and insulting the accused to be able to overwhelm him with the unpopularity thus created.

There are also other topics which arise from and against the statement of the case by the prosecution. The defence may begin by drawing arguments from the person involved, and will at times urge on general grounds that it is incredible that a father has been killed by his son or that a commander has betrayed his country to the enemy. The answer to such arguments is easy, for we may urge that bad men are capable of every crime, as is shown by every-day occurrences, or that the atrocious nature of a crime is but a poor argument against its having been committed.

At times we may base our arguments on the special circumstances of the person involved. This may be done in various ways: rank, for example, may be pleaded in defence of the accused, or at times, on the other hand, may be employed to prove his guilt on the ground that he trusted to his rank to secure impunity. Similarly poverty, humble rank, wealth may be used as arguments for or against the accused according to the talent of the advocate.

Upright character, however, and the blamelessness of his past life are always of the utmost assistance to the accused. If no charge is made against his character, counsel for the defence will lay great stress on this fact, while the accuser will attempt to restrict the judge to the sole consideration of the actual issue which the court has to decide, and will say that there must always be a first step in crime and that a first offence is not to be regarded as the occasion for celebrating a feast in honour of the defendant's character.

So much for the methods of reply which will be employed by the prosecution. But he will also in his opening speech endeavour to dispose the judges to believe that it is not so much that he is unable, as that he is unwilling to bring any charge against the character of the accused. Consequently it is better to abstain from casting any slur on the past life of the accused than to attack him with slight or frivolous charges which are manifestly false, since such a proceeding discredits the rest of our argument. Further, the advocate who brings no charges against the accused may be believed to have omitted all reference to past offences on the ground that such reference was not necessary, while the advocate who heaps up baseless charges thereby admits that his only argument is to be found in the past life of the accused, and that he has deliberately preferred to risk defeat on this point rather than say nothing at all about it.

As regards the other arguments derived from character, I have already discussed them in connexion with places of argument. The next type of proof is derived from causes or motives, such as anger, hatred, fear, greed or hope, since all motives can be classified as species of one or other of these. If any of these motives can be plausibly alleged against the accused, it is the duty of the accuser to make it appear that such motives may lead a man to commit any crime, and to exaggerate the particular motives which he selects for the purpose of his argument.

If no such motive can be alleged, he must take refuge in suggesting that there must have been some hidden motive, or in asserting that, if he committed the act, all enquiry into motive is irrelevant or that a motiveless crime is even more abominable than one which has a motive. Counsel for the defence, on the other hand, will, wherever it be possible, emphasise the point that it is incredible that any act should be committed without a motive. Cicero develops this point with great energy in a number of his speeches, but more especially in his defence of Varenus, who had everything else against him and was as a matter of fact condemned.

But if the prosecution do allege some motive, he will either say that the motive alleged is false or inadequate or unknown to the accused. For it is possible that a man may be quite ignorant of motives imputed to him. He may not, for example, have known whether the man whom he is accused of having killed had appointed him his heir or intended to prosecute him. All else failing, we may urge that motives are not necessarily of importance. For what man is there who is not liable to the emotions of fear, hatred or hope, and yet numbers of persons act on these motives without committing crime?

Nor should we neglect the point that all motives do not apply to all persons. For example, although poverty may in certain cases be a motive for theft, it will not have the same force with men such as Curius or Fabricius.

The question has been raised as to whether we should deal first with persons or motives, and different orators have given different answers: Cicero as a rule prefers to treat motives first. For my own part, if the circumstances of the case do not point strongly in either direction, I consider that the most natural course is to begin by dealing with persons. For the enquiry whether any crime can credibly be imputed to such a man as the accused is of a more general character than the question whether some one particular crime can be imputed, and in addition involves a greater correctness in division.

Still, in many cases expediency may make it desirable to reverse the order. Further, we have to seek not merely motives affecting the will, but also causes leading to error, such as drunkenness or ignorance. For just as such considerations lessen the guilt of a crime when regarded from the point of view of its quality, so they tell heavily against the criminal as far as the question of fact is concerned.

I should imagine that there could never be a case, or at any rate an actual case in the courts, in which neither side said anything about the character of the persons involved; but this is not true of causes and motives, for it is often wholly unnecessary to trouble ourselves about them, as, for example, in cases of theft or adultery, crimes which carry their motives on the face of them.

Next we must consider the intention, a term which involves a number of questions, such as whether it is probable that the accused hoped that he would be able to carry such a crime into effect, or that it would escape detection when committed, or that, even if detected, it would be pardoned or punished but lightly or after considerable lapse of time, or that the inconvenience involved by the penalty would be outweighed by the pleasure resulting from the crime: or again, whether it was worth while incurring the penalty;

and further, whether he could have done the deed at some other time and in some other way, or with greater ease or security, as Cicero says in the pro Milone, where he mentions the numerous occasions when his client could have killed Clodius with impunity. There is also the question why the accused should have chosen that particular place or time or means for the commission of the crime, a topic to which Cicero gives a thorough treatment in the same passage;

or whether, without having the least reason for the deed, he was carried away by the impulse of the moment and acted without deliberate purpose (for it is a common saying that crimes are irrational), or finally, whether he was led astray by the fact that crime had become a habit with him. Having dealt with the question whether he intended to commit the crime, we proceed to the question whether he was in a position to commit it. Under this head we discuss the place and occasion of the offence. For instance, in a case of theft we ask whether it was committed in a secluded or frequented spot, in the daytime, when witnesses are more numerous, or by night, when success is more difficult.

Consequently we shall consider all the circumstances rendering the act difficult or easy of accomplishment; these require no illustration, being numerous and familiar. This second topic is of such importance that, if it is impossible to give it satisfactory treatment, the case falls to the ground; if, however, we succeed in dealing with it adequately, we must proceed to consider whether the accused actually committed the act. But this topic involves conjecture as to intention, for it is from these facts that we infer whether he hoped for success or not. Therefore we must also consider the question of the means at his disposal, such, for example, as the retinues of Clodius and Milo.

The question whether he actually did the deed belongs, in the first place, to the second division of time, namely the present, and secondly to time that is almost, though not actually contemporary: under this latter head come circumstances such as noise, cries or groans, while concealment, fear and the like belong to subsequent time. To these must be added indications, which we have already discussed elsewhere, and words and acts antecedent or subsequent to the crime.

These words and acts are either our own or those of others. With regard to words, our own do us greater harm and bring us less profit than do those of others, while those of others bring us greater profit and do us less harm than our own. On the other hand, with regard to deeds, it is sometimes from our own and sometimes from those of others that we derive the greatest advantage, as, for example, when our opponent has done something which tells in our favour: but our own acts are always more injurious to our case than are those of others.

Again, with regard to words, we must distinguish between those whose meaning is clear and those whose significance is doubtful. The latter will necessarily give less assistance to either side, be they our own words or another's. On the other hand, any ambiguity in our own words will always tell against us, as, for example, in the following controversial theme. A son, on being asked where his father was, replied: 'Wherever he is, he is alive.' But the father was found dead at the bottom of a well.

When the ambiguity lies in words used by others, they can never do us any harm, unless he who uttered them be unknown or dead; I will give two illustrations of my meaning: A cry was heard at night, 'Be on your guard against the establishment of a tyranny'; and, on being asked who had poisoned him, he replied: 'It is not fit that you should know.' For if the speaker is available for examination, he will clear up the ambiguity.

Finally, whereas our own words and deeds can only be justified by their intention, the deeds and words of others can be disposed of in a number of different ways. My remarks on this subject have, I think, been confined to one very important class of conjectural cases, but something of what I have said will apply to all cases. For example, in cases concerned with theft, deposits and loans, arguments are derived both from possibilities (as when we enquire whether there was any money to deposit), and from persons, as when we raise the question whether it is credible that anyone deposited money with this man or trusted him with a loan, or that the claimant is bringing a false accusation, or that the accused repudiates his debt or is a thief.

But even in the case of an accusation of theft, just as in an accusation of murder, we enquire both into the act and the author, while in cases concerned with loans and deposits there are also two questions (though these are always distinct from one another), namely, whether the money was delivered and whether it has been repaid. Cases of adultery are marked by the following peculiarity, that, as a rule, the safety of two persons is involved, and it is necessary to say something of the past life of both, although some have raised the question whether both parties should be defended together. The line to be taken must depend on the circumstances of the individual case: if the defence of one party lends support to the defence of the other, I should defend them conjointly; if the reverse is the case, I should treat the two cases separately.

However, that no one may think me somewhat hasty in saying that two persons are as a rule involved in charges of adultery, I would point out that I would not assert that this is always the case. The woman alone may be accused of adultery with a person unknown: we may say, Gifts were found in the house, and money from some unknown source, and love-letters whose destination cannot be ascertained.

The case is similar in accusations of forgery: for either there are several accused or only one. The writer of a document always regards it as necessary to support the signatory, but the signatory does not always support the writer of the document, for it is always possible that he has been deceived on the matter. On the other hand, the man who is said to have called in their services, and for whom the document is alleged to have been written, will always defend both writer and signatories. The arguments employed in cases of treason or attempted tyranny will be drawn from the same sources.

But the custom prevalent in the schools of regarding everything not definitely stated in the theme as being in the speaker's favour, is likely to prove harmful to students destined for practice in the courts. You bring a charge of adultery. Who is your witness? who is your informer? You charge me with treason. What was my reward? who was my accomplice? You charge me with poisoning. Where did I buy the poison, and from whom? When did I buy it, what was the price, and whom did I employ to administer it? Or in defence of one charged with attempting to establish himself as tyrant, the declaimer will cry, Where are my weapons, and what bodyguards have I ever collected?

I do not deny that these questions should be asked, or that we should use them as far as is permitted by the rôle which we have assumed; for even in the courts I feel that it will be desirable to put such questions, if my opponent is not in a position to reply effectively; but we have often felt the lack of such freedom in the courts, whereas in the schools there is scarcely a case where one or more examples of this method are not to be found.

Similar to this is the practice which some declaimers allow themselves in their perorations of assigning children, parents and nurses to their characters at will, though it is more reasonable to call for evidence which is not explicitly mentioned in the statement of the theme than to introduce it ourselves. With regard to the method to be followed when we enquire into intention, I have said enough in dividing the subject into three questions, namely, whether the accused intended to do the deed, whether he was in a position to do it and whether he actually did it. For the method of enquiring into the purpose with which an act was committed is identical with that employed in enquiring whether the deed was intended, since it amounts to asking whether a criminal act was intended.

Further, the order in which the facts are stated may either contribute to or detract from the credibility of our case; for consistency and the reverse depend to a very great extent on the way the circumstances are arranged. But we shall be unable to detect these qualities unless we consider the circumstances in connexion with the case as a whole. None the less, it will always be necessary to consider what are best suited to be placed together.

III. Conjecture is followed by definition. For the man who cannot assert that he has done nothing, must needs take refuge in the assertion that lie has not committed the act which is alleged against him. Consequently the laws which govern definition are for the most part the same as those which govern conjecture, the only difference lying in the method to be employed in defence in cases such as those concerned with theft, deposits or adultery. For just as we say, I have not committed theft, I never received a deposit, I am not guilty of adultery, so we say, This is not theft, this is not a deposit, this is not adultery.

Sometimes we may pass from quality to definition, as in actions concerned with lunacy, cruelty and offences against the State. In such cases if it is impossible to assert that the acts alleged were right, we are left with such pleas as, To use bad language to one's wife does not amount to cruelty. Definition is the statement of the fact called in question in appropriate, clear and concise language.

As I have already said, it consists mainly in the statement of genus, species, difference and property. For example, if you wish to define a horse (for I will take a familiar example), the genus is animal, the species mortal, the diffrence irrational (since man also is mortal) and the properly neighing. Definition is employed by the orator for a number of different reasons.

For sometimes, though there may be no doubt as to a term, there is a question as to what it includes, or, on the other hand, there may be no doubt about the thing, but no agreement as to the term to be applied to it. When the term is agreed, but the thing doubtful, conjecture may sometimes come into play, as, for instance, in the question, What is god?

For the man who denies that god is a spirit permeating all things, assuredly asserts that the epithet divine is falsely applied to his nature, like Epicurus, who gives him a human form and makes him reside in the intermundane space. While both use the same term god, both have to employ conjecture to decide which of the two meanings is consistent with fact.

Sometimes again we have recourse to quality, as in the question, What is rhetoric? Is it the power to persuade or the science of speaking well? This form of question is of frequent occurrence in the courts. For instance, the question may arise whether a man caught in a brothel with another man's wife is an adulterer. Here there is no doubt about the name; it is the significance of the act which is in doubt, since the question is whether he has committed any sin at all. For if he has sinned, his sin can only be adultery.

There is a different type of question where the dispute is concerned with the term to be applied, which depends on the letter of the law: it is a form of question which can only arise in the courts from the actual words on which the dispute turns. Take as examples the questions, whether suicide is a form of homicide, or whether the man who forces a tyrant to kill himself can be considered a tyrannicide, or whether magical incantations are equivalent to the crime of poisoning. In all these cases there is no doubt about the facts, for it is well known that there is a difference between killing oneself and killing another, between slaying a tyrant and forcing him to suicide, between employing incantations and administering a deadly draught, but we enquire whether we are justified in calling them by the same name.

Though I hardly like to differ from Cicero, who follows many authorities in saying that definition is concerned with identity and difference (since he who denies the applicability of one term must always produce another term which he regards as preferable), for my own part I consider that definition falls into three types, which I may perhaps call species.

For at times it is convenient merely to enquire whether one particular term is applicable to a given thing, as in the question whether an act committed in a brothel is adultery. If we deny that it is adultery, there is no need to state what it should be called, since we have given a total denial to the charge. Secondly, there are occasions when the question is which of two terms is to be applied to a thing, as in the question whether an act is theft or sacrilege. It may be sufficient for the defence that it is not sacrilege, but it is still necessary to show what else it is, and therefore we must define both.

Thirdly, there are times when the question concerns things which are different in species, and we ask whether two different things are to be called by the same name, in spite of the fact that each has a special name of its own: for example, is tile same name applicable both to a love-potion and a poison? But in all disputes of this kind the question is whether one thing is another thing as well, since the name in doubt does without question apply to something else. It is sacrilege to steal a sacred object from a temple; is it also sacrilege to steal private property from a temple? It is adultery to lie with another man's wile in one's own house; is it adultery to do so in a brothel? It is tyrannicide to slay a tyrant; is it tyrannicide to force him to slay himself?

Consequently the syllogism, to which I shall return later, is virtually a weaker form of definition, since while definition seeks to determine whether one thing has the same name as another, syllogism seeks to determine whether one thing is to be regarded as identical with another.

There is moreover great variety in definitions. For instance, persons will give different verbal expression to things about which they are really in agreement: thus rhetoric is defined as the science of speaking well, as the science of correct conception or correct expression of what we have to say, as the science of speading in accordance with the excellence of an orator and again of speaking to the purpose. And we must take care to discover how it is that definitions, identical in meaning, differ in the form in which they are expressed. However, this is a subject for discussion and not for a quarrel.

Definition is sometimes required to explain rare or obscure words such as clarigatio or erctum citum, or again to explain familiar words such as penus or litus. This variety in definition has caused some writers to include it under conjecture, others under quality and others again under legal questions.

Some, on the other hand, entirely reject the elaborate and formal methods of reasoning employed by dialectic, regarding such ingenuity as suited rather to quibbles over words in philosophical discussions than as likely to carry much weight in the performance of the duties of an orator. For though in dialogue definition may serve to fetter the person who has got to reply in chains of his own making, or may force him to silence, or even to reluctant confession of a point which tells against himself, it is of less use in forensic cases.

For there we have to persuade the judge, who, even though he may be tied and bound with our words, will still dissent in silence, unless he is brought really into touch with the actual facts. And what need has a pleader for such precision of definition? Even if I do not say that man is an animal, mortal and rational, surely I shall still be able, by setting forth the numerous properties of his body and mind in more general terms, to distinguish him from gods or dumb beasts.

Again, may not the same thing be defined in more than one way, as Cicero does when he says, What do we mean when we say' commonly ': surely we mean 'by all men'? May it not be given a wide and varied treatment such as is frequently employed by all orators? For it is rare to find orators falling victims to that form of slavery introduced from the practice of the philosophers and tying themselves down to certain definite words; indeed it is absolutely forbidden by Marcus Antonius in the de Oratore of Cicero. For it is a most dangerous practice, since, if we make a mistake in a single word, we are like to lose our whole case, and consequently the compromise adopted by Cicero in the pro Caecina is the safest course to follow; this consists in setting forth the facts without running any risks over the exactness of our terminology. These are his words: Judges, the violence which threatens our lives and persons is not the only kind of violence: there is a much more serious form which by the threat of death fills our minds with panic and often turns them from their natural condition of stability.

Or again, we may prove before we define, as Cicero does in the Philippics, where he proves that Servius Sulpicius was killed by Antony and introduces his definition at the conclusion in the following terms:— For assuredly the murderer was he who was the cause of his death. I would not, however, deny that such rules should be employed, if it will help our case, and that, if we can produce a definition which is at once strong and concise, it will be not merely an ornament to our speech, but will also produce the strongest impression, provided always that it cannot be overthrown.

The order to be followed in definition is invariable. We first ask what a thing is, and then, whether it is this. And there is generally more difficulty in the establishment than in the application of a definition. In determining what a thing is, there are two things which require to be done: we must establish our own definition and destroy that of our opponent. Consequently in the schools, where we ourselves imagine our opponent's reply, we have to introduce two definitions, which should suit the respective sides of the case as well as it is in our power to make them. But in the courts we must give careful consideration to the question whether our definition may not be superfluous and irrelevant or ambiguous or inconsistent or even of no less service to our opponents than to ourselves, since it will be the fault of the pleader if any of these errors occur. On the other hand, we shall ensure the right definition, if we first make up our minds what it is precisely that we desire to effect: for, this done, we shall be able to suit our words to serve our purpose. To make my meaning clearer, I will follow my usual practice and quote a familiar example. A man who has stolen private money from a temple is accused of sacrilege.

There is no doubt about his guilt; the question is whether the name given by the law applies to the charge. It is therefore debated whether the act constitutes sacrilege. The accuser employs this term on the ground that the money was stolen from a temple: the accused denies that the act is sacrilege, on the ground that the money stolen was private property, but admits that it is theft. The prosecutor will therefore give the following definitions, It is sacrilege to steal anything from a sacred place. The accused will reply with another definition, It is sacrilege to steal something sacred. Each impugns the other's definition.

A definition may be overthrown on two grounds: it may be false or it may be too narrow. There is indeed a possible third ground, namely irrelevance, but this is a fault which no one save a fool will commit.

[It is a false definition if you say, A horse is a rational animal, for though the horse is an animal, it is irrational. Again, a thing which is common to something else cannot be a property of the thing defined.] In the case under discussion, then, the accused alleges that the definition given by the accuser is false, whereas the accuser cannot do the same by his opponent's definition, since to steal a sacred object is undoubtedly sacrilege. He therefore alleges that the definition is too narrow and requires the addition of the words or from a sacred place.

But the most effective method of establishing and refuting definitions is derived from the examination of properties and differences, and sometimes even from considerations of etymology, while all these considerations will, like others, find further support in equity and occasionally in conjecture. Etymology is rarely of assistance, but the following will provide an example of its use. For what else is a 'tumult' but a disturbance of such violence as to give rise to abnormal alarm? And the name itself is derived from this fact.

Great ingenuity may be exercised with regard to properties and differences, as for instance in the question whether a person assigned to his creditor for debt, who is condemned by the law to remain in a state of servitude until he has paid his debt, is actually a slave. One party will advance the following definition, A slave is one who is legally in a state of servitude. The other will produce the definition, A slave is one who is in a state of servitude on the same terms as a slave (or, to use the older phrase, 'who serves as a slave'). This definition, though it differs considerably from the other, will be quite useless unless it is supported by properties and differences.

For the opponent will say that the person in question is actually serving as a slave or is legally in a state of servitude. We must therefore look for properties and differences, to which in passing I devoted a brief discussion in my fifth book. A slave when manumitted becomes a freedman: a man who is assigned for debt becomes a free man on the restoration of his liberty. A slave cannot acquire his freedom without the consent of his master: a man assigned for debt can acquire it by paying his debt without the consent of his master being necessary. A slave is outside the law; a man assigned for debt is under the law. Turning to properties, we may note the following which are possessed by none save the free, the three names (praenomen, nomen and cognomen) and membership of a tribe, all of which are possessed by the man assigned for debt.

By settling what a thing is we have come near to determining its identity, for our purpose is to produce a definition that is applicable to our case. Now the most important element in a definition is provided by quality, as, for example, in the question whether love be a form of madness. To this point in our procedure belong those proofs which according to Cicero are peculiar to definition, that is, proofs drawn from antecedents, consequents, adjuncts, contraries, causes, effects and similarities, with the nature of which I have already dealt.

I will, however, quote a passage from the pro Caecina in which Cicero includes brief proofs drawn from origins, causes, effects, antecedents and consequents: Why then did they fly? Because they were afraid. What were they afraid of? Obviously of violence. Can you then deny the beginning, when you have admitted the end? But he also argued from similarity: Shall not that which is called violence in war be called violence in peace as well Arguments may also be drawn from contraries, as for instance in the question whether a love-potion can be a poison, in view of the fact that a poison is not a love-potion. In order that my young students (and I call them mine, because the young student is always dear to me) may form a clearer conception of this second kind of definition, I will once more quote a fictitious controversial theme.

Some young men who were in the habit of making merry together decided to dine on the sea-shore. One of their party failed to put in an appearance, and they raised a tomb to him and inscribed his name thereon. His father on his return from overseas chanced to land at this point of the shore, read the name and hung himself. It is alleged that the youths were the cause of his death.

The definition produced by the accuser will run as follows: The man whose act leads to another's death is the cause of his death. The definition given by the accused will be, He who wittingly commits an act which must necessarily lead to another's death, is the cause of his death. Without any formal definition it would be sufficient for the accuser to argue as follows: You were the cause of his death, for it was your act that led to his death: but for your act he would still be alive.

To which the accused might answer, It does not necessarily follow that the man whose act leads to another's death should be condemned forthwith. Were this so, the accuser, witnesses and judges in a capital case would all be liable to condemnation. Nor is the cause of death always a guilty cause. Take for instance the case of a man who persuades another to go on a journey or sends for his friend from overseas, with the result that the latter perishes in a shipwreck, or again the case of a man who invites another to dine, with the result that the guest dies of indigestion. Nor is the act of the young men to be regarded as the sole cause of death. The credulity of the old man and his inability to bear the shock of grief were contributory causes. Finally, had lie been wiser or made of sterner stuff, he would still be alive. Moreover the young men acted without the least thought of doing harm, and the father might have suspected from the position of the tomb and the traces of haste in its construction that it was not a genuine tomb. What ground then is there for condemning them, for everything else that constitutes homicide is lacking save only the contributory act?

Sometimes we have a settled definition on which both parties are agreed, as in the following example from Cicero: Majesty resides in the dignity of the Roman power and the Roman people. The question however, is, whether that majesty has been impaired, as for example in the ease of Cornelius. But even although the case may seem to turn on definition, the point for decision is one of quality, since there is no doubt about the definition, and must be assigned to the qualitative basis. It is a mere accident that I have come to mention quality at this moment, but in point of fact quality is the matter that comes next in order for discussion.

IV. In speaking of quality we sometimes use the word in its most general sense, which covers a number of different questions. For we enquire sometimes into the nature and form of things: as for instance whether the soul is immortal or whether god is to be conceived of in human form. Sometimes, on the other hand, the question turns on size and number, as, for instance, what is the size of the sun or whether there are more worlds than one. In all these cases we arrive at our conclusions by conjecture, yet each involves a question of quality.

Such questions are sometimes treated in deliberative themes: for example, if Caesar is deliberating whether to attack Britain, he must enquire into the nature of the Ocean, consider whether Britain is an island (a fact not then ascertained), and estimate its size and the number of troops which lie will require for the invasion. Under the same head of quality fall questions whether certain things should be done or not and certain objects sought or avoided: such topics are specially adapted for deliberative themes, but occur with some frequency in controversial themes as well, the only difference being that in the latter we deal with what is past and in the former with the future.

Similarly all the topics of demonstrative oratory involve a qualitative basis. The facts are admitted, and the question turns on their quality, the dispute being entirely concerned with rewards or penalties or their quantity. The case is therefore of two kinds, simple or comparative, the former dealing with what is just, the latter with what is juster, or most just. When the point for decision is the penalty to be inflicted, the duty of the pleader will be to defend, extenuate or excuse the act on which the charge is based, or even, according to some, to plead for mercy.

By far the strongest line that can be taken in defence is to assert that the act which forms the subject of the charge is actually honourable. A man is disinherited because he went on military service, stood for office or married without his father's consent. We defend this act. This form of defence is called κατ᾽ ἀντίληψιν by the followers of Hermagoras, that is, defence by objection, the term being used with reference to the purport of the defendant's plea. I can find no exact Latin translation of the term; we call it an absolute defence. But in such cases the question is concerned with the justice or injustice of the act alone.

Justice is either natural or conventional. Natural justice is found in actions of inherent worth.

Under this head come the virtues of piety, loyalty, self-control and the like. To these some add the rendering of like for like. But this view must not be adopted without consideration: for to retaliate, or meet violence with violence on the one hand, does not imply injustice on the part of the aggressor, while on the other hand it does not follow that the first act was just merely because the two acts were alike. In cases where there is justice on both sides, the two parties must both come under the same law and the same conditions, and it would not perhaps be untrue to say that things can never be spoken of as like if there is any point in which they are dissimilar. Convention, on the other hand, is to be found in laws, customs, legal precedents and agreements.

There is another form of defence by which we defend an act in itself indefensible by arguments drawn from without. This the Greeks call κατ᾽ ἀντίθεσιν by opposition. Here again there is no Latin equivalent, since we call it defence by assumption.

The strongest line to take in this form of defence is to defend the act forming the subject of the charge by appealing to its motive. An example of this is provided by the defence put forward on behalf of Orestes, Horatius or Milo. The term ἀντέγκλημα, or counter-accusation, is employed when our defence consists entirely in accusing the person whom our opponents are seeking to vindicate. He was killed, but he was a robber; he was blinded, but he was a ravisher.

There is another form of defence based on an appeal to the motives of the act which is the opposite of that which I have just described It consists not in defending the act per se, as we do when we employ the absolute defence, nor in opposing another act to it, but in appealing to the interests of the State, of a number of persons, of our opponent himself or finally at times of ourselves, provided always that the act in question is such as we might lawfully do in our own interests. If, however, the accuser is a stranger and insists on the letter of the law, this form of defence will invariably be useless, though it may serve our turn if the dispute is of a domestic character.

For example, in a suit concerned with the question of disinheritance a father may, without reflecting on himself, say to his sons that his act was of importance to his own interests, and the same plea may be urged by a husband accused of cruelty by his wile or a son who alleges that his father is insane. But in such cases the position of the man who seeks to avoid loss is stronger than that of him who aims at positive advantage.

Precisely similar methods are also employed in questions that occur in real life. For the scholastic themes concerned with the disowning of children are on exactly the same footing as the cases of sons disinherited by their parents which are tried in the public courts, or of those claims for the recovery of property which are tried in the centum viral court: themes dealing with cruelty find an actual parallel in those cases in which the wife claims the restoration of her dowry, and the question is whose fault it was that led to the divorce: and again the theme where the son accuses his father of madness has its analogy in cases where a suit is brought for the appointment of a guardian.

Under the same heading as the appeal to public or personal interest comes the plea that the act in question prevented the occurrence of something worse. For in a comparison of evils the lesser evil must be regarded as a positive good: for example, Mancinus may defend the treaty made with the Numantines on the ground that it saved the army from annihilation. This form of defence is called ἀντίστασις by the Greeks, while we style it defence by comparison.

Such are the methods by which we may defend an act. If it is impossible to defend an act either on its merits or with the assistance of arguments from without, the next best course will be to shift the charge, if possible, to another. It is for this reason that the basis of competence has been held to apply even to those who cannot plead the letter of the law in this connexion. In some cases, then, the blame will be thrown on a person: for example, Gracchus, when accused of making the treaty with the Numantines (and it was fear of this accusation that seems to have led him to bring forward the democratic laws of his tribuneship) may plead that he made it as the representative of his commander-in-chief.

At times, on the other hand, the blame may be shifted to some thing: for instance, a person who has failed to comply with some testamentary injunction may plead that the laws forbade such compliance. The Greek term for such shifting is μετάστασις. If these methods of defence are out of the question, we must take refuge in making excuses. We may plead ignorance. For example, if a man has branded a runaway slave and the latter is subsequently adjudged to be a free man, he may deny that he was cognisant of the truth. Or we may plead necessity; for instance, if a soldier overstays his leave, he may plead that his return was delayed by floods or ill health.

Again, the blame is often cast upon fortune, while sometimes we assert that, although we undoubtedly did wrong, we did so with the best intentions. Instances of these two latter forms of excuse are, however, so numerous and obvious that there is no need for me to cite them here. If all the above-mentioned resources prove unavailable, we must see whether it may not be possible to extenuate the offence. It is here that what some call the quantitative basis comes into play.

But when quantity is considered in reference to punishment or reward, it is determined by the quality of the act, and therefore in my opinion comes under the qualitative basis, as also does quantity which is used with reference to number by the Greeks, who distinguish between ποσότης and πηλικότης: we, however, have only one name for the two.

In the last resort we may plead for mercy, a though most writers deny that this is ever admissible in the courts. Indeed Cicero himself seems to support this view in his defence of Quintus Ligarius where he says, I have pleaded many causes, Caesar, some of them even in association with yourself, so long as your political ambitions prevented you from abandoning the bar, but never have I pleaded in words such as these, ' Forgive him, gentlemen, he erred, he made a slip, he did not think that it mattered, he will never do it again,' and so on.

On the other hand, in addressing the senate, the people, the emperor or any other authority who is in a position to show clemency, such pleas for mercy have a legitimate place. In such cases there are three points based on the circumstances of the accused which are most effective. The first is drawn from his previous life, if he has been blameless in his conduct and deserved well of the state, or if there is good hope that his conduct will be blameless for the future and likely to be of some use to his fellow men; the second is operative if it appears that he has been sufficiently punished already on the ground that he has suffered other misfortunes, or that his present peril is extreme, or that he has repented of his sin; while thirdly we may base his appeal on his external circumstances, his birth, his rank, his connexions, his friendships.

It is, however, on the judge that we shall pin our highest hopes, if the circumstances be such that acquittal will result in giving him a reputation for clemency rather than for regrettable weakness. But even in the ordinary courts appeals for mercy are frequently employed to a large extent, although they will not colour the whole of our pleading. For the following form of division is common:— Even if he had committed the offence, he would have deserved forgiveness, a plea which has often turned the balance in doubtful cases, while practically all perorations contain such appeals.

Sometimes indeed the whole case may rest on such considerations. For example, if a father has made an express declaration that he has disinherited his son because lie was in love with a woman of the town, will not the whole question turn on the point whether it was the father's duty to pardon such an offence and whether it is the duty of the centumviral court to overlook it? Nay, even in penal prosecutions governed by strict forms of law we raise two separate questions: first whether the penalty has been incurred, and secondly whether, if so, it ought to be inflicted. Still the view of the authorities to whom I have referred that an accused person cannot be saved from the clutches of the law by this method of defence is perfectly correct.

With regard to rewards, there are two questions which confront us: namely, whether the claimant is deserving of any reward, and, if so, whether he deserves so great a reward. If there are two claimants, we have to decide which is the more worthy of the two; if there are a number, who is the most worthy. The treatment of these questions turns on the kind of merit possessed by the claimants. And we must consider not merely the act (whether it has merely to be stated or has to be compared with the acts of others), but the person of the claimant as well. For it makes a great difference whether a tyrannicide be young or old, man or woman, a stranger or a connexion.

The place may also be discussed in a number of ways: was the city in which the tyrant was killed one inured to tyranny or one which had always been free? was he killed in the citadel or in his own house? The means, too, and the time call for consideration: was he killed by poison or the sword? was he killed in time of peace or war, when he was intending to lay aside his power or to venture on some fresh crime?

Further, in considering the question of merit, the danger and difficulty of the act will carry great weight, while with regard to liberality it will similarly be of importance to consider the character of the person from whom it proceeds. For liberality is more pleasing in a poor man than in a rich, in one who confers than in one who returns a benefit, in a father than in a childless man. Again, we must consider the immediate object of the gift, the occasion and the intention, that is to say, whether it was given in the hope of subsequent profit; and so on with a number of similar considerations. The question of quality therefore makes the highest demands on the resources of oratory, since it affords the utmost scope for a display of talent on either side, while there is no topic in which the emotional appeal is so effective.

For conjecture has often to introduce proofs from without and uses arguments drawn from the actual subject matter, whereas the real task of eloquence is to demonstrate quality: there lies its kingdom, there its power, and there its unique victory. Verginius includes under quality cases concerned with disinheritance, lunacy, cruelty to a wife, and claims of female orphans to marry relatives. The questions thus involved are, it is true, frequently questions of quality, while some writers style them questions of moral obligation.

But the laws governing these cases sometimes admit of other bases. For example, conjecture is involved in quite a number of such cases, as when the accused urges either that he did not commit the act or, if he did, acted with the best intentions. I could quote many examples of this kind. Again, it is definition which tells us what precisely is meant by lunacy or cruelty to a wife. For as a rule the laws cited in such themes involve certain legal questions, though not to such an extent as to determine the basis of the case.

But this notwithstanding, if the actual fact cannot be defended, we may in the last resort base our defence on legal grounds, in which case we shall consider how many and what cases there are in which a father may not disinherit his son, what charges fail to justify an action for cruelty, and under what circumstances a son is not allowed to accuse his father of lunacy.

Disinheritance may be of two kinds. In the first case it is for a completed crime: for example, the son who is disinherited may be a ravisher or an adulterer: in the second case it is for a crime which is still incomplete and terminable; an instance of this will be the case where the son is disinherited because of disobedience to his father. The first form of disinheritance always demands a certain harshness when the father pleads his case, since the act is irrevocable, whereas in the latter his pleading will be of a kindly and almost persuasive nature, since he would prefer not to disinherit him. On the other hand, the pleading of the sons should in both cases be of a subdued character and couched in a conciliatory tone.

I know that those who delight in making covert attacks upon the father under the disguise of some figure of speech will disagree with me: and I would not deny that their procedure may sometimes be justifiable, since the theme may conceivably be such as to demand it; but it is certainly to be avoided wherever possible. However, I shall deal with the whole question of figures in a later book.

The treatment of the theme of cruelty to a wife is not unlike that of the theme of disinheritance; for both demand a certain moderation on the part of the accuser. Cases concerned with lunacy arise either out of what has been done or out of something which may or may not be done in the future.

In the former case the pleader is free to attack as he will, but must none the less do so in such a manner that, while denouncing the act, he will yet express pity for the father on the ground that he has erred by reason of his infirmity. On the other hand, in the latter case, where the act has not yet taken place and there is nothing to prevent the father changing his purpose, he must begin by a prolonged attempt to induce him to change his mind, and then, and only then, complain that it is madness and not depravity of character that prevents him from listening to the voice of reason; and the more he praises his past character, the easier will it be to prove the change which it has undergone owing to the inroads of the disease.

The accused, wherever possible, must assume a temperate tone in his defence, for the reason that as a rule anger and excitement are near akin to madness. All these cases have this much in common, that the accused does not always defend his act, but often pleads excuse and asks for pardon. For these are domestic quarrels, in which the fact that the offence is an isolated case, due to error and of a less serious character than alleged, will sometimes suffice to secure an acquittal.

There are, however, a number of other controversial themes involving quality, as, for example, cases of assault. In these, although at times the accused denies that he committed the assault, the pleading as a rule is concerned with fact and intention.

Then there are cases concerned with the appointment of a prosecutor, which are known as divinations. In this connexion Cicero, who was indicting Verres on the instruction of our Sicilian allies, adopts the following division-to the effect that the main point for consideration is, by whom those the redress of whose wrongs forms the subject of the trial would prefer to be represented, and by whom the accused would least desire them to be represented.

But in the great majority of cases the questions raised are, which claimant has the strongest motives for undertaking the role of accuser, which is likely to bring the greatest energy or talent to the task, and which is likely to press the charge with the greatest sincerity.

Next we may take cases concerned with guardianship, in which it is usual to enquire whether it is necessary to investigate anything save the accounts, and whether anything can be demanded of the guardian except the honest execution of his trust; his sagacity and the success of his administration being beside the mark. Cases of fraud on the part of an agent, which are styled cases of conduct of business when they occur in the actual courts, are of a similar nature, since they also are concerned with the administration of a trust.

In addition to these we have the fictitious cases of the schools which deal with crimes not covered by the law, where the question is as a rule either whether the crime is really not covered by the law or whether it is a crime, though on rare occasions both questions are raised. Cases of misconduct on the part of an ambassador are of frequent occurrence among the Greeks, even in actual life: in these the legal question is raised whether it is lawful to deviate at all from one's instructions and for how long the accused was technically an ambassador, since in some cases the ambassador's duty is to convey a communication to a foreign power and in others to bring one back. Take for example the case of Heius, who gave evidence against Verres after performing his duties as ambassador. But in such cases the most important question turns on the nature of the deed complained of Next come cases of action contrary to the interests of the state. In these we meet with legal quibbles as to what is the meaning of action contrary to the interests of the state, and whether the action of the accused was injurious or profitable, or whether the interests of the state suffered at his hands or merely on his account: but the most important question is that of fact. There are also cases of ingratitude; in these we raise the question whether the accused has really received any kindness. It is only rarely that the fact can be denied, as denial is in itself a sign of ingratitude. But there are the further questions as to the extent of the kindness and whether it has been repaid.

If it has not been repaid, does this necessarily involve ingratitude? Was it in his power to repay? Did he really owe the return which was demanded of him? What is his intention? Somewhat simpler are cases of unjust divorce, a form of controversy which has this peculiarity, that the accuser defends and the defendant accuses.

Further there are cases where a senator sets forth to the senate the reasons which determine him to commit suicide, in which there is one legal question, namely, whether a man who desires to kill himself in order to escape the clutches of the law ought to be prevented from so doing, while the remaining questions are all concerned with quality. There are also fictitious cases concerned with wills, in which the only question raised is one of quality, as, for instance, in the controversial theme quoted above, where the philosopher, physician and orator all claim the fourth share which their father had left to the most worthy of his sons. The same is true of cases where suitors of equal rank claim the hand of an orphan and the question confronting her relatives is which is the most suitable.

I do not, however, intend to discuss every possible theme, since fresh ones can always be invented, nor yet to deal with all the questions to which they give rise, since these vary with circumstances. But I cannot help expressing my astonishment that Flavus, for whose authority I have the highest respect, restricted the range of quality to such an extent in the text-book which he composed for the special guidance of the schools.

Quantity also, as I have already stated, falls as a rule, though not always, under the head of quality, whether it is concerned with measure or number. Measure, however, sometimes consists in the valuation of a deed with a view to determining the amount of guilt or the amount of benefit involved, while, on the other hand, it sometimes turns on a point of law, when the dispute is under what law a man is to be punished or rewarded.

For example is a ravisher to pay 10,000 sesterces because that is the penalty appointed by law, or is he liable to capital punishment as a murderer because his victim hanged himself? In such cases those who plead as if there were a question between two laws, are wrong: for there is no dispute about the fine of 10,000, since it is not claimed by the prosecution.

The point on which judgment has to be delivered is whether the accused is guilty of causing his victim's death. The same type of case will also bring conjecture into play, when, for example, the question in dispute is whether the accused shall be punished with banishment for life or for five years. For the question then is whether he caused his death willingly or not.

Again, there are questions concerned with numerical quantity which turn on a point of law, such as the questions whether thirty rewards are due to Thrasybulus, or whether, when two thieves have stolen a sum of money, they are each to be required to refund fourfold or twofold. But in these cases, too, valuation of the act is necessary, and yet the point of law also turns on quality.

V. He who neither denies nor defends his act nor asserts that it was of a different nature from that alleged, must take his stand on some point of law that tells in his favour, a form of defence which generally turns on the legality of the action brought against him.

This question is not, however, as some have held, always raised before the commencement of the trial, like the elaborate deliberations of the praetor when there is a doubt as to whether the prosecutor has any legal standing, but frequently comes up during the course of the actual trial. Such discussions fall into two classes, according as the point in dispute arises from an argument advanced by the prosecution or from some prescription

(or demurrer) put forward by the defence. There have indeed been some writers who have held that there is a special prescriptive basis; but prescription is covered by precisely the same questions that cover all other laws.

When the dispute turns on prescription, there is no need to enquire into the facts of the case itself. For example, a son puts forward a demurrer against his father on the ground that his father has forfeited his civil rights. The only point which has to be decided is whether the demurrer can stand. Still, wherever possible, we should attempt to create a favourable impression in the judge as to the facts of the case as well, since, if this be done, he will be all the more disposed to give an indulgent hearing to our point of law: for example, in actions taking the form of a wager and arising out of interdicts, even though the question is concerned solely with actual possession, the question as to tile right to possession not being raised, it will be desirable to prove not merely that the property was actually in our possession, but that it was ours to possess.

On tile other hand, the question more frequently turns on intention. Take the law Let a hero choose what reward he will. I deny that he is entitled to receive whatever he chooses. I cannot put forward any formal demurrer, but none the less I use the intention as against the letter of the law just as I should use a demurrer. In both cases the basis is the same.

Moreover every law either gives or takes away punishes or commands, forbids or permits, and involves a dispute either on its own account or on account of another law, while the question which it involves will turn either on the letter or the intention. The letter is either clear or obscure or ambiguous.

And what I say with reference to laws will apply equally to wills, agreements, contracts and every form of document; nay, it will apply even to verbal agreements. And since I have classified such cases under four questions or bases, I will deal with each in turn.

Lawyers frequently raise the question of the letter and the intention of the law, in fact a large proportion of legal disputes turn on these points. We need not therefore be surprised that such questions occur in the schools as well, where they are often invented with this special purpose. One form of this kind of question is found in cases where tile enquiry turns both on the letter and the spirit of a law.

Such questions arise when the law presents some obscurity. Under these circumstances both parties will seek to establish their own interpretation of the passage and to overthrow that advanced by their opponent. Take for example the following case. A thief shall refund four times the amount of his theft. Two thieves have jointly stolen 10,000 sesterces. 40,000 are claimed from each. They claim that they are liable only to pay 20,000 each. The accuser will urge that the sum which he claims is fourfold the amount stolen; the accused will urge that the sum which they offer to pay is fourfold. The intention of the law will be pleaded by both parties.

On the other hand, the dispute may turn on a passage of the law which is clear in one sense and doubtful in another. The son of a harlot shall not address the people. A woman who had a son became a prostitute. The youth is forbidden to address the people. Here there is no doubt about the son of one who was a prostitute before his birth, but it is doubtful whether the law applies to the case of one born before his mother became a prostitute.

Another question which is not infrequently raised is as to the interpretation of the law forbidding an action to be brought twice on the same dispute, the problem being whether the word twice refers to the prosecutor or the prosecution. Such are the points arising out of the obscurity of the law. A second form of question turns on some passage where the meaning is clear. Those who have given exclusive attention to this class of question call it the basis concerned with the obvious expression of the law and its intention. In such circumstances one party will rest their case on the letter, the other on the intention of the law.

There are three different methods in which we may combat the letter. The first comes into play where it is clear that it is impossible always to observe the letter of the law. Children shall support their parents under penalty of imprisonment. It is clear, in the first place, that this cannot apply to an infant. At this point we shall turn to other possible exceptions and distinguish as follows. Does this apply to everyone who refuses to support his parent? Has this particular individual incurred the penalty by this particular act?

The second arises in scholastic themes where no argument can be drawn from the particular law, but the question is concerned solely with the subject of the dispute. A foreigner who goes up on to the wall shall be liable to capital punishment. The enemy had scaled the wall and were driven back by a foreigner. His punishment is demanded.

In this case we shall not have two separate questions, namely, whether every foreigner who goes up on the wall is liable to the penalty, and whether this particular foreigner is liable, since no more forcible argument can be brought against the application of the letter of the law than the fact in dispute, but the only question to be raised will be whether a foreigner may not go on to the wall even for the purpose of saving the city. Therefore we shall rest our case on equity and the intention of the law. It is, however, sometimes possible to draw examples from other laws to show that we cannot always stand by the letter, as Cicero did in his defence of Caecina.

The third method becomes operative when we find something in the actual words of the law which enables us to prove that the intention of the legislator was different. The following theme will provide an example. Anyone who is caught at night with steel in his hands shall be thrown into prison. A man is found wearing a steel ring, and is imprisoned by the magistrate. In this case the use of the word caught is sufficient proof that the word steel was only intended by the law in the sense of a weapon of offence.

But just as the advocate who rests his case on the intention of the law must wherever possible impugn the letter of the law, so he who defends the letter of the law must also seek to gain support from the intention. Again, in cases concerned with wills it sometimes happens that the intention of the testator is clear, though it has not been expressed in writing: an example of this occurs in the trial of Curius, which gave rise to the well-known argument between Lucius Crassus and Scaevola.

A second heir had been appointed in the event of a posthumous son dying while a minor. No posthumous son was born. The next of kin claimed the property. Who could doubt that the intention of the testator was that the same man should inherit in the event of the son not being born who would have inherited in the event of his death? But he had not written this in his will.

Again, the opposite case, that is to say, when what is written is obviously contrary to tile intention of the writer, occurred quite recently. A man who had made a bequest of 5000 sesterces, on altering his will erased the word sesterces and inserted pounds of silver. But it was clear that he had meant not 5000 but 5 pounds of silver, because the weight of silver mentioned in the bequest was unparalleled and incredible.

The same basis includes such general questions as to whether we should stand by the letter or the intention of the document, and what was the purpose of the writer, while for the treatment of such questions we must have recourse to quality or conjecture, with which I think I have dealt in sufficient detail.

The next subject which comes up for discussion is that of contrary laws.

For all writers of text-books are agreed that in such cases there are two bases involving the letter and the intention of the law respectively. This view is justified by the fact that, when one law contradicts another, both parties attack the letter and raise the question of intention, while the point in dispute, as regards each law, is whether we should be guided by it at all.

But it is clear to everybody that one law cannot contradict another in principle (since if there were two different principles, one law would cancel the other), and that the laws in question are brought into collision purely by the accidents of chance. When two laws clash, they may be of a similar nature, as for instance if we have to compare two cases in which a tyrannicide and a brave man are given the choice of their reward, both being granted the privilege of choosing whatever they desire. In such a case we compare the deserts of the claimants, the occasions of the respective acts and the nature of the rewards claimed.

Or the same law may be in conflict with itself, as in the case where we have two brave men, two tyrannicides or two ravished women, when the question must turn either on time (that is, whose claim has priority) or on quality (that is, whose claim is the more just). Again, we may have a conflict between diverse, similar or dissimilar laws.

Diverse laws are those against which arguments may be brought without reference to any contradictory law. The following theme will provide an example. A magistrate shall not quit the citadel. One who has rendered heroic service to his country may choose what reward he pleases. A magistrate who left his post and saved his country, demands an amnesty for his conduct. In this case, even though there be no other law covering the case, we may raise the question whether a hero ought to be granted anything he chooses to claim. Again, many conclusive arguments may be brought against the letter of the law restricting the movements of the magistrate: for example, a fire may have broken out in the citadel, or a sally against the enemy may have been necessary.

Laws are styled similar when nothing can be opposed to one except the other. Tyrannicides shall have their statues set up in the gymnasium. A statue of a woman shall not be set up in the gymnasium. A woman killed a tyrant. Here are two conflicting laws: for a woman's statue cannot under any other circumstances be erected in the gymnasium, while there is no other circumstance which can bar the erection of the statue of a tyrannicide in the gymnasium.

Laws are styled dissimilar when many arguments can be urged against one, while the only point which can be urged against the other is the actual subject of dispute. An example is provided by the case in which a brave man demands the pardon of a deserter as his reward. For there are many arguments, as I have shown above, which can be urged against the law permitting a hero to choose whatever reward he will, but the letter of the law dealing with the crime of desertion cannot be overthrown under any circumstances save the choice of rewards to which I have just referred.

Again, the point of law is either admitted by both parties or disputed. If it be admitted, the questions which are raised will as a rule be such as the following. Which of the two laws is the most stringent? Does it concern gods or men, the state or private individuals, reward or punishment, great things or small? Does it permit, forbid or command?

Another common question is which of the two laws is the oldest; but the most important question is which of the two laws will suffer less by its contravention, as for example in the case of the hero and the deserter just mentioned, in which case, if the deserter is not put to death, the whole law is ignored, whereas, if he be put to death, the hero will still have another choice left open to him. It is, however, of the utmost importance to consider which course is best from the point of view of morality and justice, a problem for the solution of which no general rules can be laid down, as it will depend on the special circumstances of the case.

If, on the other hand, the point of law is disputed, either one party or both in turn will argue the point. Take the following case as an example. A father shall be empowered to arrest his son, and a patron to arrest his freedman. Freedmen shall be transferred to their patron's heir. A certain man appointed the son of a freedman as his heir. The son of the freedman and the freedman himself both claim the right to arrest the other. Here the father claims his right over the son, while the son, in virtue of his new position as patron, denies that his father possessed the rights of a father, because he was in the power of his patron.

Laws containing two provisions may conflict with themselves in exactly the same way as two laws may conflict. The following will serve as an illustration. The bastard born before a legitimate son shall rank as legitimate, the bastard born after the legitimate son shall only rank as a citizen. All that I have said about laws will also apply to decrees of the senate. If decrees of the senate conflict with one another or with the laws, the basis will be the same as if laws only were concerned.

The syllogistic basis has some resemblance to the basis concerned with the letter and intention of the law, since whenever it comes into play, one party rests his case on the letter: there is, however, this difference between the two bases, that in the latter we argue against the letter, in the present beyond the letter, while in the latter the party defending the letter aims at securing that in any case the letter may be carried into effect, whereas in the present his aim will be to prevent anything except the letter being carried into effect. The syllogism is sometimes employed in conjunction with definition: for often if the definition be weak it takes refuge in the syllogism. Assume a law to run as follows: A woman who is a poisoner shall be liable to capital punishment. A wife gave her husband a love-potion to cure him of his habit of beating her. She also divorced him. On being asked by her relatives to return to him, she refused. The husband hung himself. The woman is accused of poisoning. The strongest line for the accuser to take will be to assert that the love-potion was a poison. This involves definition. If it proves weak, we shall have recourse to the syllogism, to which we shall proceed after virtually dropping our previous argument, and which we shall employ to decide the question whether she does not deserve to be punished for administering the lovepotion no less than if she had caused her husband's death by poison.

The syllogistic basis, then, deduces from the letter of the law that which is uncertain; and since this conclusion is arrived at by reason, the basis is called ratiocinative. It may be subdivided into the following species of question. If it is right to do a thing once, is it right to do it often? Example: A priestess found guilty of unchastity is thrown from the Tarpeian rock and survives. It is demanded that she shall be thrown down again. If the law grants a privilege with reference to one thing, does it grant it with reference to a number? Example: A man kills two tyrants together and claims two rewards.

If a thing is legal before a certain occurrence, is it legal after it? Example: The ravisher took refuge in flight. His victim married. The ravisher returned and the woman demands to be allowed her choice. Is that which is lawful with regard to the whole, lawful with regard to a part? Example: It is forbidden to accept a plough as security. He accepted a ploughshare. Is that which is lawful with regard to a part, lawful with regard to the whole? Example: It is forbidden to export wool from Tarentum: he exported sheep.

In all these cases the syllogism rests on the letter of the law as well: for the accuser urges that the provisions of the law are precise. He will say, I demand that the priestess who has broken her vows be cast down: it is the law, or The ravished woman demands the exercise of the choice permitted her by law, or Wool grows on sheep, and so on.

But to this we may reply, The law does not prescribe that the condemned woman should be thrown down twice, that the ravished woman should exercise her choice under all circumstances, that the tyrannicide should receive two rewards, while it makes no mention of ploughshares or of sheep. Thus we infer what is doubtful from what is certain. It is a more difficult task to deduce from the letter of the law that which is not actually prescribed by the letter, and to argue because that is the case, so also is this. Take the following problems. The man who kills his father shall be sewn up in a sack. He killed his mother, or It is illegal to drag a man from his own house into the court. He dragged him from his tent.

Under this heading come questions such as the following: if there is not a special law applicable to the case, ought we to have recourse to an analogous law? is the point in question similar to what is contained in the letter of the law? Now it should be noted that what is similar may be greater, equal or less. In the first ease we enquire whether the provisions of the law are sufficient, or, if they are insufficient, whether we should have recourse to this other law. In both cases it is a question of the intention of the legislator. But the most effective form of treatment in such cases will be to appeal to equity.

I turn to tile discussion of ambiguity, which will be found to have countless species: indeed, in the opinion of certain philosophers, there is not a single word which has not a diversity of meanings. There are, however, very few genera, since ambiguity must occur either in a single word or in a group of words.

Single words give rise to error, when the same noun applies to a number of things or persons (the Greeks call this homonymy): for example, it is uncertain with regard to the word gallus whether it means a cock or a Gaul or a proper name or an emasculated priest of tybele; while Ajax may refer either to the son of Telamon or the son of Oileus. Again, verbs likewise may have different meanings, as, for example, cerno.

This ambiguity crops up in many ways, and gives rise to disputes, mole especially in connexion with wills, when two men of the same name claim their freedom or, it may be, an inheritance, or again, when the enquiry turns on the precise nature of the bequest.

There is another form of ambiguity where a word has one meaning when entire and another when divided, as, for example, ingenua, armameniam or Corvinum. The disputes arising from such ambiguities are no more than childish quibbles, but nevertheless the Greeks are in the habit of making them the subject for controversial themes, as, for example, in the notorious case of the αὐλητρίς, when the question is whether it is a hall which has fallen down three times (αὔλη τρίς) or a flute-player who fell down that is to be sold.

A third form of ambiguity is caused by the use of compound words; for example, if a man orders his body to be buried in a cultivated spot, and should direct, as is often done, a considerable space of land surrounding his tomb to be taken from the land left to his heirs with a view to preserving his ashes from outrage, an occasion for dispute may be afforded by the question whether the words mean in a cultivated place (in culto loco) or in an uncultivated place (inculto loco).

Thus arises the Greek theme about Leon and Pantaleon, who go to law because the handwriting of a will makes it uncertain whether the testator has left all his property to Leon or his property to Pantaleon. Groups of words give rise to more serious ambiguity. Such ambiguity may arise from doubt as to a case, as in the following passage: —

I say that you, O prince of Aeacus' line,

Rome can o'erthrow.

Or it may arise from the arrangement of the words, which makes it doubtful what the exact reference of some word or words may be, more especially when there is a word in the middle of the sentence which may be referred either to what precedes or what follows, as in the line of Virgil which describes Troilus as lora tenens tamen, where it may be disputed whether the poet means that he is still holding the reins, or that, although he holds the reins, he is still dragged along.

The controversial theme, A certain man in his will ordered his heirs to erect statuam auream hastam tenentem,' turns on a similar ambiguity; for it raises the question whether it is the statue holding the spear which is to be of gold, or whether the spear should be of gold and the statue of some other material. The same result is even more frequently produced by a mistaken inflexion of the voice, as in the line: quinquaginta uhi erant centum inde occidit Achilles.

It is also often doubtful to which of two antecedents a phrase is to be referred. Hence we get such controversial themes as, My heir shall be bound to give my wife a hundred pounds of silver according to choice, where it is left uncertain which of the two is to make the choice. But in these examples of ambiguity, the first may be remedied by a change of case, the second by separating 1 the words or altering their position, the third by some addition.

Ambiguity resulting from the use of two accusatives may be removed by the substitution of the ablative: for example, Lachetem audivi percussisse Demeam (I heard that Demea struck Laches, or that L. struck D.) may be rendered clear by writing a Lachete percussum Demeam (that D. was struck by L.). There is, however, a natural ambiguity in the ablative case itself, as I pointed out in the first book. For example, caelo decurrit aperto leaves it doubtful whether the poet means he hastened down through the open sky, or when the sky was opened for him to pass.

Words may be separated by a breathing space or pause. We may, for instance, say statuam, and then, after a slight pause, add auream hastam, or the pause may come between statuam auream and haslam. The addition referred to above would take the form quod elegerit ipse, where ipse will show that the reference to the heir, or quod elegerit ipsa, making the reference to the wife. In cases where ambiguity is caused by the addition of a word, the difficulty may be eliminated by the removal of a word, as in the sentence nos flentes illos deprehendimus.

Where it is doubtful to what a word or phrase refers, and the word or phrase itself is ambiguous, we shall have to alter several words, as, for example, in the sentence, My heir shall be bound to give him all his own property, where his own is ambiguous. Cicero commits the same fault when he says of Gaius Fannius, He following the instructions of his father-in-law, for whom, because he had not been elected to the college of augurs, he had no great affection, especially as he had given Quintus Scaevola, the younger of his sons-in-law, the preference over himself.. For over himself may refer either to his father-in-law or to Fannius.

Again, another source of ambiguity arises from leaving it doubtful in a written document whether a syllable is long or short. Cato, for example, means one thing in the nominative when its second syllable is short, and another in the dative or ablative when the same syllable is long. There are also a number of other forms of ambiguity which it is unnecessary for me to describe at length.

Further, it is quite unimportant how ambiguity arises or how it is remedied. For it is clear in all cases that two interpretations are possible, and as far as the written or spoken word is concerned, it is equally important for both parties. It is therefore a perfectly futile rule which directs us to endeavour, in connexion with this basis, to turn the word in question to suit our own purpose, since, if this is feasible, there is no ambiguity.

In cases of ambiguity the only questions which confront us will be, sometimes which of the two interpretations is most natural, and always which interpretation is most equitable, and what was the intention of the person who wrote or uttered the words. I have, however, given sufficient instructions in the course of my remarks on conjecture and quality, as to the method of treating such questions, whether by the prosecution or the defence.

There is, however, a certain affinity between all these bases. For in definition we enquire into the meaning of a term, and in the syllogism, which is closely connected with definition, we consider what was the meaning of the writer, while it is obvious that in the case of contrary laws there are two bases, one concerned with the letter, and the other with the intention. Again, definition is in itself a kind of ambiguity, since it brings out two meanings in the same term.

The basis concerned with the letter and the intention of the law involves a legal question as regards the interpretation of the words, which is identical with the question arising out of contrary laws. Consequently some writers have asserted that all these bases may be resolved into those concerned with the letter and intention, while others hold that in all cases where the letter and the intention of a document have to be considered, it is ambiguity that gives rise to the question at issue. But all these bases are really distinct, for an obscure point of law is not the same as an ambiguous point of law. Definition, then, involves a general question as to the actual nature of a term, a question which may conceivably have no connexion whatsoever with the content of the case in point. In investigations as to the letter and the intention, the dispute turns on the provisions contained in the law, whereas the syllogism deals with that which is not contained in the law. In disputes arising out of ambiguity we are led from the ambiguous phrase to its conflicting meanings, whereas in the case of contrary laws the fight starts from the conflict of their provisions.

The distinction between these bases has therefore been rightly accepted by the most learned of rhetoricians, and is still adopted by the majority and the wisest of the teachers of to-day. It has not been possible in this connexion to give instructions which will cover the arrangement to be adopted in every case, though I have been able to give some.

There are other details concerning which I can give no instructions without a statement of the particular case on which the orator has to speak. For not only must the whole case be analysed into its component topics and questions, but these subdivisions themselves require to be arranged in the order which is appropriate to them. For example, in the exordium each part has its own special place, first, second and third, etc., while each question and topic requires to be suitably arranged, and the same is true even of isolated general questions.

For it will not, I imagine, be represented that sufficient skill in division is possessed by the man who, after resolving a controversial theme into questions such as the following, whether a hero is to be granted any reward that he may claim, whether he is allowed to claim private property, whether he may demand any woman in marriage, whether he may claim to marry a woman who already possesses a husband, or this particular woman, then, although it is his duty to deal with the first question first, proceeds to deal with them indiscriminately as each may happen to occur to him, and ignores the fact that the first point which should be discussed is whether we should stand by the letter or the intention of the law, and fails to follow the natural order, which demands that after beginning with this question he should then proceed to introduce the subsidiary questions, thereby making the structure of his speech as regular as that of the human body, of which, for example, the hand is a part, while the fingers are parts of the hand, and the joints of the fingers. It is precisely this method of arrangement which it is impossible to demonstrate except with reference to some definite and specific case.

But it is clearly useless to take one or two cases, or even a hundred or a thousand, since their number is infinite. It is the duty of the teacher to demonstrate daily in one kind of case after another what is the natural order and connexion of the parts, so that little by little his pupils may gain the experience which will enable them to deal with other cases of the same character. For it is quite impossible to teach everything that can be accomplished by art.

For example, what painter has ever been taught to reproduce everything in nature? But once he has acquired the general principles of imitation, he will be able to copy whatever is given him. What vase-maker is there who has not succeeded in producing a vase of a type which he had never previously seen?

There are, however, some things which depend not on the teacher, but on the learner. For example, a physician will teach what treatment should be adopted for different diseases, what the dangers are against which he must be on his guard, and what the symptoms by which they may be recognised. But he will not be able to communicate to his pupil the gift of feeling the pulse, or appreciating the variations of colour, breathing and temperature: this will depend on the talent of the individual. Therefore, in most instances, we must rely on ourselves, and must study cases with the utmost care, never forgetting that men discovered our art before ever they proceeded to teach it.

For the most effective, and what is justly styled most economical arrangement of a case as a whole, is that which cannot be determined except when we have the specific facts before us. It consists in the power to determine when the exordium is necessary and when it should be omitted; when we should make our statement of facts continuous, and when we should subdivide it; when we should begin at the very beginning, when, like Homer, start at the middle or the end;

when we should omit the statement of facts altogether; when we should begin by dealing with the arguments advanced by our opponents, and when with our own; when we should place the strongest proofs first and when the weakest; in what cases we should prefix questions to the exordium, and what preparation is necessary to pave the way for these questions; what arguments the judge will accept at once, and to what he requires to be led by degrees; whether we should refute our opponent's arguments as a whole or in detail; whether we should reserve emotional appeals for the peroration or distribute them throughout the whole speech; whether we should speak first of law or of equity; whether we should first advance (or refute) charges as to past offences or the charges connected with the actual trial;

or, again, if the case is complicated, what order we should adopt, what evidence or documents of any kind should be read out in the course of our speech, and what reserved for a later stage. This gift of arrangement is to oratory what generalship is to war. The skilled commander will know how to distribute his forces for battle, what troops he should keep back to garrison forts or guard cities, to secure supplies, or guard communications, and what dispositions to make by land and by sea.

But to possess this gift, our orator will require all the resources of nature, learning and industrious study. Therefore let no man hope that he can acquire eloquence merely by the labour of others. He must burn the midnight oil, persevere to the end and grow pale with study: he must form his own powers, his own experience, his own methods: he must not require to hunt for his weapons, but must have them ready for immediate use, as though they were born with him and not derived from the instruction of others.

The road may be pointed out, but our speed must be our own. Art has done enough in publishing the resources of eloquence, it is for us to know how to use them.

And it is not enough merely to arrange the various parts: each several part has its own internal economy, according to which one thought will come first, another second, another third, while we must struggle not merely to place these thoughts in their proper order, but to link them together and give them such cohesion that there will be no trace of any suture: they must form a body, not a congeries of limbs.

This end will be attained if we note what best suits each position, and take care that the words which we place together are such as will not clash, but will mutually harmonise. Thus different facts will not seem like perfect strangers thrust into uncongenial company from distant places, but will be united with what precedes and follows by an intimate bond of union, with the result that our speech will give the impression not merely of having been put together, but of natural continuity. I fear, however, that I have been lured on from one thing to another and have advanced somewhat too far, since I find myself gliding from the subject of arrangement to the discussion of the general rules of style, which will form the opening theme of the next book.