ICO 



NA TURE 



[Dec. 6, 1877 



little volume. We think the idea of making such a col- 

 lection a happy one, not only for scholastic purposes, but 

 also for the use of those who wish to be able at any time 

 easily to refer to any of the passages in Latin authors in 

 which our island is referred to. Mr. Cayzer gives also 

 translations of some of the chief references in Greek 

 writers. We should think, if teachers and examiners 

 could be persuaded to break through custom, the intro- 

 duction of such a book into schools would add interest to 

 the reading of Latin, and furnish, besides, the little fellows 

 with a stock of valuable information. Most of the cuts 

 are appropriate, several being old friends. 



LETTERS TO THE EDITOR 



\Tht Editor does not hold himself responsible for opinions expressed 

 by his correspondents. Neither can he undei-take to return^ 

 or to correspond with the writers of, rejected manuscripts. 

 No notice is taken of anonymous communuations. 



TTie Editor urgently requests correspondents to hap their letters as 

 short as possible. The pressure on his space is so great that it 

 is impossible otherwise to ensure the appearance even of com- 

 munications containing interesting and novel facts.\ 



The Colour-Sense of the Greeks 



Mr. Gladstone has shown that the language of Homer is an 

 inadequate vehicle for conveying precise and nicely distinguished 

 ideas of colour. Whether the nation that was content to describe 

 colours so imperfectly was also incapable of subtle perception of 

 tones of colour is clearly another question. Language does not 

 keep pace with perception unless a practical or resthetic necessity 

 arises for expressing what is perceived in words to other people. 



Practical necessity gives names to pigments and bright objects, 

 such as flowers and precious stones, rather than to tones of 

 colour ; the aesthetic necessity that lies upon the artist to utter 

 what he has felt will naturally lead to imitative expression sooner 

 than to an expression that is merely symbolical. In other words 

 an early race will learn to use colour with nicety for decorative 

 and pictorial purposes before it develops the distinctions of 

 language requisite for accurate word-painting. 



That this was actually the case among the Greeks appears, I 

 think, very clearly in a passage of Ion which is preserved to us 

 in Athenseus Deipnos., Lib. xiii. cap. 81 (p. 603 seq.). Ion, who 

 was a contemporary of Sophocles, describes an evening which 

 he spent with the great tragedian in Chios. Sophocles, admiring 

 the blushing face of a little boy who served the wine, quoted, 

 with high approval, a line of Phrynicus : — 



" The light of love gleams on the purplecheek." 



On this a certain pedantic grammarian breaks in — "In sooth, 

 Sophocles, thou art skilled in poetry ; but yet Phrynicus spoke 

 not well when he called the cheeks of a beautiful person purple. 

 For if a portrait-painter were to colour the cheeks of this boy 

 with purple pigment he would no longer appear beautiful. It is 

 not fitting to compare what is beautiful with what is not so." 

 Sophocles laughs at the objection, and replies — " Neither, then, 

 my fiiend, wilt thou be pleased with that line of Simonides 

 which, to the Greeks, has appeared very well said : — 



* The maiden sending forth her voice from her purple mouth ; ' 



nor with the poet, when he says, ' golden-haired Apollo ; ' 

 for if the painter made the hair of the god golden and not 

 black, his picture would be less excellent. Nor wilt thou be pleased 

 with him [Homer] who said * rosy-fingered,' for if one were to 

 dip the fingers in rose-colour, one would produce the hands, not 

 of a fair woman, but of a dyer of purple." This retort produced 

 a general laugh, and confounded the pedant not a little. 



The Greeks, then, were perfectly aware of the insufficiency of 

 the poetic vocabulary of colour ; and accordingly they did not 

 expect descriptive rendering of colour from the poet. This, it 

 is plain, is a circumstance that must constantly be kept in 

 view in any attempt to find in the poetry of the Greeks a 

 measure of the development of their colour-sense. 



Aberdeen, December 3 W. Robertson Smith 



The Comparative Richness of Faunas and Floras 

 Tested Numerically 



In his letter in Nature, vol. xvii. p. 9, Prof. Newton has 

 strongly brought out the abstudity of comparing districts of very 



different areas by the proportionate number of species to area in 

 each. On this principle he shows that to be equally rich with the 

 small island of Rodriguez, Madagascar ought to possess four times 

 as many species of birds as exist throughout the whole world ! 

 It does not, however, by any means follow that the method thus 

 expo-ed may not be of value in comparing regions of approxi- 

 mately equal area, as is the case with several of the primary 

 regions, to determine the comparative richness of which Mr. 

 Sclater first applied it. I have not Mr. Sclater's paper at hand, 

 but it is my impression that he made no attempt to sho w— " that 

 the proper mode of comparing the wealth or poverty of one 

 fauna with another was to state the proportion which the number 

 of species composing it bears to the area over which they range " — 

 as Prof. Newton implies that he did, but that he merely adopted 

 this method as the only one readily available for the comparison 

 of his regions. Although I took the opportunity of making 

 some corrections in the figures, I never committed myself to the 

 principle ; and I very soon afterwards found that it was not to be 

 trusted. As, however, several later writers have made use of it 

 without remark, it will be interesting to consider where the exact 

 point of the fallacy lies, and with what modifications the method 

 can be trusted to give useful and consistent results. 



If we compare two islands of almost exactly equal areas, such 

 as Ceylon and Tasmania, and find that the one has twice or 

 three times as many species of mammals or birds as the other, 

 it will be generally admitted that we express the fact correctly 

 when we say that, as regards such a group of animals, the one is 

 twice or thrice as rich as the other ; and the same may be said 

 of two countries or two continents of identical areas. For on the 

 supposition that there is a general correspondence between the 

 numbers of rare and common, of local and of wide-spread 

 species in the two areas compared (and this seems probable), 

 then the average number of distinct species to be m^t with on 

 one spot, or to be seen during a journey of equal length, will 

 be proportionate to the total number of species in the two 

 areas. But now let us divide one of the two continents 

 or islands which we are comparing into two or more parts. We 

 know, as a matter of fact, that one-half the area will always 

 contain much more than half the total number of species, while 

 one-tenth of the area will contain immensely more than one- tenth 

 of the species. To take an example : the county of Sussex is 

 about one-eightieth part the area of the British Isles, yet it ac- 

 tually contains full two-thirds of the total number of flowerino' 

 plants, both being estimated by the same flora (Babington's. 

 "Manual," fifth edition, British Isles 1,536 species, Sussex 

 1,059 species). If we now compare either Britain or Sussex with 

 an eqtial area on the continent of Europe or North America, we 

 may obtain an instructive estimate of the comparative richness of 

 their respective floras ; but if we compare unequal areas, and 

 then endeavour to equalise them by getting the proportions of 

 species to area, we shall obtain erroneous results, which will 

 become literally absurd when the areas compared are very 

 unequal. 



The problem remains, how to compare unequal areas of which 

 we possess the zoological or botanical statistics. We can only 

 do so by equalising them, and this may not be so difficult as at 

 first sight appears. For example, let us take the Palsearctic and 

 North American regions, in which the species of birds are nearly 

 equal in number, but the areas are as about seven to three. The 

 number of the Palasarctic species have, however, been propor- 

 tionately increased of late years, and if we take the western half 

 of the Palaearctic region so as to include North Africa and Persia 

 we shall have an area about equal to the Nearctic region, and a 

 number of species perhaps one-sixth or one-eighth les>, which 

 will thus represent the comparative richness ol these two areas. 

 The eastern half of the region, including Japan and North China, 

 is probably as rich as the western ; while the iijtc:rmediate portion 

 is poorer in species. Combining these three portions, and taking 

 the average, we should perhaps find the Palasarctic region about 

 four-fifths or five-sixths as ricti as the Nearctic, instead of less 

 thaa one-half, as shown by the method of proportionate areas. 



Whenever we know how xaz.riy peculiar species any district 

 contains, we can deduct its area from the total area of the region 

 to be compared, and this numbtr of peculiar species, from the 

 fauna of the region ; and by this means we may reduce two 

 unequal regions to comparative equality. Again, all detached 

 portions or inlands should be omitted in estimating the compara- 

 tive richness of regions, because they affect these regions very 

 unequally. By adding Britam to Europe you increase the area 

 without adding to the fauna, and thus make the region seem 

 poorer ; while by adding Madagascar to Africa, or New Zealand 



