^{page no 15 note 1} The Thesaurus distinguishes, besides the metaphorical uses and that of ‘mesh’ in a net, two classes of meanings. They are (I) nota deformans, (2) nota latiore sensu, i.e. spots of animals, marks in wood, etc. (1) is subdivided into (a) maculae per se natae, as freckles, moles, etc., (b) maculae extrinsecus aspergine, contagione effectae. The earliest example of (I) (a) is from Plautus, while (2) is not represented earlier than Varro and his contemporaries; but this may well be due to accident. A definite case of (1)(b), macula meaning a spot or blot made by sprinkling something, i.e. the use which according to Dr. Palmer is the original one, is not to be found earlier than Publilius Syrus and Ovid. It may be said that Plautus implies it in his use of maculosus, Bacch. 434: ‘fieret corium tam maculosum quam est nutricis pallium’ (O.S.). But this is none too certain. The schoolboy's skin became maculosum because he had been soundly thrashed, which certainly is not sprinkling, and it is doubtful whether the nurse was wearing a garment spotted by the baby's imperfect table manners or simply woven in a variegated pattern (H. J. R.). The metaphorical macula in the love-sick breast of Agorastocles, , Poen. 198Google Scholar, which could not washed out without injuring the fabric of that tender organ, is quite as good a figure if taken from natural spots as if it refers to a blot or stain. In any case, if we are not mistaken, the fairly common use of macula to signify the mesh of a net is in favour of our view, for it is hard to see how that can be compared to a stain or other mark made by sprinkling or the like, easy to see its resemblance to the spots which form part of the texture of some creatures’ hides.

We might be asked why, if this is so, colloquial Latin has no proper word for a stain. Though it may be outside the limits of our obligation to reply to such a question, we suggest in passing that a word did exist, namely labes, whether that is originally different from labes ‘ruin’ or merely another sense of the same word (see Ernout–Meillet, s.v.). In either case, the competition of the other word, or sense, supported as it was by the verb labi, may have ousted it from daily speech, causing macula to take its place. For homonymity as a cause for the replacement of words, see Palmer's excellent account of this phenomenon, Intr. to Modern Linguistics, pp. 109 ff.

^{page no 16 note 1} Despite semasiological appearances it would seem impossible to identify Sanscrit mahati with *magit. It is true that the n of manh (mah) is not necessarily part of the root; but the aspirate is of a different character from that of mahā or aham, which, an innovation peculiar to Sanscrit, corresponds to the unaspirated media of magnus and ego in the other Indo-European languages; whereas the aspirate of mah is also found in Avestan and therefore most likely of IndoEuropean origin. Avestan mimayzo further shows that, if *mago were to be combined with mahati, it would have to be separated from magnus, since, as opposed to the final prepalatal of maz- (= mag-nus), the verb has final guttural.

^{page no 16 note 2} We readily agree that our examples were not well chosen; that left-handed is beside the point, and that barbatus belongs to a different category. Yet, the suffix to makes adjectival derivatives both from nominal and adverbial stems and from roots (there is no such thing as an ‘adjectival root’). That in the latter circumstance this formation should be restricted to participles of full-blown verbs is surely an exaggerated statement. Whether, in order to connect multus with melior, it is necessary to have recourse to a verb *melo (which happens to be represented in Greek μέλω) may seem doubtful; less doubtful that in e.g. vas-tus and va-nus (from vas-nus) we have a close parallel to mac-tus and mag-nus, and yet there is no primary verb to which vastus could be attached. However that may be, we must insist that it does not affect the main issue: even if *mago were not attested in Old Irish its reconstruction would present no methodical difficulty.

^{page no 17 note 1} The offering not accepted by the god is referred to with ille, the present offering with hic; e.g. 141 quod tibi illuc porco neque satisfactum est, te hoc porco piaculo.

^{page no 19 note 1} One point which might be made is that Cicero, a man of considerable learning and a friend of Varro, might very well take his ideas of the meaning of maclare from Varro, who as we have seen probably connected it with magnus. If this is so, he cannot be appealed to by either side as a witness to the true and original meaning of the word, unless, that is, Varro's etymology was substantially right.