misunderstanding of his authority. See for instance my notes on the account of Thales § 25, of Anaximenes § 26, of Parmenides § 28, of Xenophon § 31.

The Epicurean exposition, contained in §§ 43-56, is far superior to the historical section, but it suffers from curtailment, just where full explanation was most needed. Unhappily Cicero had not time to think out a difficulty; so when he comes to one, he either omits, or satisfies himself with a rendering which is unintelligible to himself as well as to every one else; see especially what is said of the divine images in § 49 compared with SS 105 and 109. In fairness it must however be allowed that he is writing for Roman readers and has to select or reject with the thought of what will be most in accordance with their taste, just as the late Dr Whewell did in his Platonic Dialogues for English Readers.

The Academic criticism which occupies the rest of the book contains much that is interesting, but, here too, flippant assertion not unfrequently takes the place of argument. Thus there is no pretence of arguing the question between a plenum and a vacuum (§ 65 foll.); the speaker dogmatically asserts his preference for the former, therefore the latter is wrong. The objections to anthropomorphism are well stated in §§ 76-102, but Cicero has either misunderstood or has confused the argument on the value of general experience, as a criterion of truth, and the possibility of a unique experience (see nn. on § 87). In SS 103, 104 Cotta announces his intention to examine the Epicurean account of the habitation and manner of life of their gods, but in § 105 hurries on to a discussion of the theory of images. Possibly this change of plan may have arisen, as Schwencke suggests, from the discovery that the original treatise from which he is translating, travelled beyond the topics introduced in the speech of Velleius. In any case it is a fault in the construction of the dialogue, and deprives us of information, which would have been very welcome, as to the nature of existence in the intermundia. The question raised in §§ 105-110 relates to the possibility of distinguishing between objective and subjective images; what right have we to assume that the phantasms of divinities are more real than those of absent or non-existent persons or things? Even if we assume their reality, what right have we to attribute happiness to beings without virtue (since without action) and without the sensual pleasures which are allotted to man? Can they even be said to be free from pain, when they are in constant

danger from the incoming and outgoing atoms? The remainder of the book is occupied in showing that the Epicurean notion of a deity, incapable of action and absorbed in his own pleasure, who has no feeling for men, and is altogether unconnected with them, is really atheistic and inconsistent with any kind of piety or holiness. It is to be noticed that the Epicurean defence (at etiam liber est Epicuri de sanctitate) meets a double rejoinder §§ 115, 123. Is this a sign that Cicero had before him two criticisms of Epicurus, one, say, by Philo (159), the other by Posidonius, or are they alternative summaries of the argument of the latter, which have both been inserted by mistake? We shall see other examples of the same sort of carelessness in the following books.

In the 2nd book we have perhaps the most important contribution to theological thought which has come down to us from classical antiquity. It wants the inspiration, the passionate earnestness of Plato, but it covers a wider range; it is a store-house in which are preserved the best achievements of Greek philosophy in this department from the time of Socrates to that of Cicero. The arrangement may be confused, many of the special theories advanced may be obsolete, many of the facts misunderstood or inaccurately stated, but the general proof here given of a rationally ordered universe, and of a providential care for man can never lose its interest or value. It holds good against all theories of evolution, whether ancient or modern, which would make mind posterior to matter. The main lines of the proof are that religious belief is natural to man; that it is confirmed by the signs of superhuman power, wisdom and goodness visible in the universe; that man cannot be the highest thing in the universe, as he would be if the universe. were irrational; rather that it is from it he derives his reason as well as the gross elements of which the body is composed; that the common source of the reason of all men must far surpass the particles of reason dispersed in individual men; that the harmony and sympathy of all the parts of the universe proves it to be under the control of one guiding spirit; that mind or soul originates all motion. Then follows the argument from the Scale of Existence: we observe the gradual ascent from vegetable to animal, from animal to man, the last showing the potentiality of virtue and wisdom, hence we infer a higher stage, the divine, which is absolutely virtuous and wise. Nature strives after perfection in all its parts; this striving cannot be frustrated; there may be partial hindrances, but

there is no external power to check the progress of the whole; therefore the universe as a whole must attain perfection.

The larger portion of the 2nd book is occupied with the subject of Providence. This is argued 1st from our idea of the Divine nature as active and benevolent, and 2nd from the skill manifested in the universe, which attests the wisdom of the great Artist, just in the same way as the orrery attests the wisdom of Archimedes. The skill of the Creator is then shown in detail, 1st as regards the earth and the heavenly bodies, 2nd as regards the adaptations visible in vegetable and animal life, 3rd and above all in the case of man. It is further shown that the universe exists for the sake of its rational inhabitants, and that all things tend to the good of man, that providential care extends even to individual men, that virtue and wisdom are divine gifts, that the philosopher is dear to God and can never experience what is really harmful.

So far I think we may be sure that Cicero would go along with Balbus. It is no more than he has repeatedly said in his own person elsewhere, except as to the Scale of Existence, to which we find resemblances, it is true, but no exact parallel in the passages quoted in my notes. There are other parts of his discourse which are less in harmony with what we know of Cicero's opinions from other treatises. Such are the identification of heat with intelligence, the ascription of life, thought and volition to the material universe and the heavenly bodies, the sanction accorded by the Stoics to the popular mythology as representing either the varied activity of the Supreme Being, the personification of abstract qualities, or the divinity of the human soul; to which we may add the belief in divination'.

When we go on to inquire into the arrangement of the 2nd book, there is much to find fault with. The main divisions are by no means clear. As is pointed out in Vol. II p. xxii, much that is placed under

1 Cicero, speaking in his own person, asserts the existence and the immateriality of God, and ascribes to him the origin of all motion and the fatherhood of the human soul Tusc. 1 66 (a quotation from his own Consolatio), animorum nulla in terris origo inveniri potest...quicquid est illud quod sentit, quod sapit, quod vivit, quod viget, caeleste et divinum ob eamque rem aeternum sit necesse est. Nec vero deus ipse...alio modo intellegi potest nisi mens soluta quaedam et libera, segregata ab omni concretione mortali, omnia sentiens et movens, ipsaque praedita motu sempiterno cf. ib. 1 30, 36, 60, 63 (the Creator is to the universe as Archimedes to his orrery), 68 foll., Leg. 1 21, 11 15 foll., Milo 83, 84, Harusp. Resp. 19.

the 1st head, would have come more naturally under the 2nd, and much that is placed under the 3rd would have come better under the 4th. In one place Cicero seems to have confused himself, and commences his 4th division out of its proper order in § 133, giving a second commencement in § 155. Then we have the superfluous Aratean section (SS 104-114), and the omission of much interesting matter at the end of the book, in reference to the calamities of the good and the difficulties alleged against the moral government of the world. For faults of detail see my notes on etenim 16, crassissima regione 17, cum alio juncta 29, absoluti operis effectum 35, ex utraque re 49, aetherios cursus 54, suis seminibus 58, vis major, regi non potest 61, dentes et pubertatem 86, where particular arguments seem to be imperfectly stated. For mistranslations of the Greek original see on obductus, cujus sub pedibus, 110, posteriore trahens 113.

It is more difficult to take a general view of the 3rd book than of the preceding, as so large a portion, probably more than one third, has been lost. It will be seen from the analysis, as well as from the Essay which follows, upon the Sources of this book, that the arrangement of what remains is again unsatisfactory. Cicero is embarrassed throughout by having to meet a later Stoic argument out of an earlier Academic treatise, in which the topics are different and differently arranged. This explains why, after Cotta has announced his intention to treat several of the arguments adduced for the Divine existence under the 3rd head, instead of under the 1st, as Balbus had done (11 17, 18), he introduces them under the 2nd head without giving any reason for his change of purpose. We will take the different arguments in order with reference to the corresponding parts of the 2nd book. There is certainly some weight in the objections urged to the argument from universal consent, viz. that the object of popular belief is not the God of the Stoics, and that it is inconsistent in those who regard the majority as fools, to attach any importance to what the majority believe (S$ 10, 11); still these objections hardly apply to the arguments as stated in Bk II. Consensus is cited there as a proof not of any special Stoic doctrine, but of the existence of a Divine Governor; and a careful distinction is made between temporary opinion and fixed belief, especially where the latter becomes stronger with the advance in civilization. The Stoic arguments derived from recorded epiphanies and the practice of divination, are fairly met by denial of the facts and questioning the utility of a knowledge of the future; the self-devotion of Decius

was prompted by policy, not by religion; to suppose otherwise would be to impute injustice to the Gods (S$ 11-15). This sets aside one of the grounds assigned for the prevalence of religious belief by Cleanthes; his 2nd ground, that of the terrible phenomena of nature, is allowed as a fact; the two others are deferred along with the arguments of Zeno and Chrysippus to the 3rd head. The 2nd branch of the discussion deals with the Divine nature. This begins in § 20 with a distinct reference to the corresponding part of the argument of Balbus (11 45). In both the question is qualis eorum natura sit; both refer to the stupefying influence of custom. Cotta then proceeds to challenge Balbus' assertion mundum animantem esse et deum, and the proof alleged for it nihil mundo esse melius. 'It no more follows from this,' he says, 'that the world must be possessed of reason than that the city of Rome is a reasoning creature, or that, if it is not, it must be reckoned of less value than the ant which is possessed of reason.' (Cf. II 45, 16.) But the same proof had been quoted as from Zeno in II 21, so Cotta recurs to that (11 22), in spite of his avowed intention of leaving it for the section on Providence, and replies that on the same principle we might argue that the world could read a book. In § 23 he deals with another argument of Zeno's given in II 22, putting it in a more general form (apparently with a reference to the Socratic argument in II 18) 'everything which exists is derived from the world, and the world can produce nothing unlike itself, therefore the human reason is a product of the world and resembles it.' 'On the same principle,' he says, 'we might maintain that the world could play the flute.' The next argument touched on by Cotta is that which deduces the divinity of the stars from their regular motions, apparently referring to II 54 foll. Similarly we might argue for the divinity of quartan fevers' (III 24). For an examination of these objections see nn. on the particular passages. In § 25 Cotta goes back to Chrysippus (11 16, 17). His 1st argument is that if there is anything in the world beyond man's power to make, he who made it must be God'; the 2nd that 'if there were no Gods, man would be the best thing in the universe, which it would be the extreme of arrogance to suppose'; the 3rd that 'the world is too beautiful to have been built simply for the habitation of man.' Cotta's answer to the 1st is that it ignores the distinction between nature and reason (which of course has no force against the Stoics who identified nature with reason, and does not in the least degree affect the inference that there is a superhuman power at work in the