302 Mr Philip Sewell on tite [sess. lii. 



Certain striking correlations also afford us definite hints as 

 to the cause of colour. Of these we may notice those cases 

 in which a change of colour in the flower or fruit (due to 

 some constitutional alteration) is indicated in an altered or 

 similar shade in the colour of the ordinary leaves. This has 

 been seen in black, red, or white fruited vines ; in Pernettya, in 

 Erica, Cineraria, Chrysanthemum, Camellia, and other plants. 

 Darwin mentions several such correlations in Variation of 

 Animah and Plants under Domestication; and it must be 

 within the knowledge of all that skilled gardeners will, with 

 remarkable precision' recognise different florists' varieties by 

 the habit of growth, or by minute differences of leaf alone, 

 and this among hundreds of the same genus. 



Quite as striking is the correlation of secreting surfaces or 

 honey glands, with more or less brilliancy of colour. 



The secreting surface is essentially a more or less localised 

 part of the plant, where not only is there no carbon-assimi- 

 lation, l)ut a surplus of somewhat altered nutritive material 

 is being given out by the plant. The much coloured and 

 viscid flower-stalk of Lychnis viscaria is also something more 

 than the outcome of a happy vaiiation with a deadly sticki- 

 ness to repel insects, and a brilliant colour to attract them ; 

 the colour and the secretion are naturally linked together. 

 So also tlie colour of honey guides, though perhaps as clearly 

 subject to selection as any part of the plant, has a natural 

 reason thus afforded for its presence, at any rate in its less 

 specialised form. 



Similarly, also, we see some reason for the presence of 

 " spots of colour in the more highly modified parts," and for 

 what has been called " compensation " in variation, as in 

 the case of Dianthus barbatus, with its crowded, brilliantly 

 colouied, transitory flowers mentioned by Dr Taylor. 



It is not possible to mention this subject of tlie correlation 

 of colour to other products of elaboration without referring 

 to a most interesting passage from an article by Mr liussel 

 Wallace, when offering objections to Mr Grant Allen's treat- 

 ment of sexual selection. He says that, among certain 

 groups of butterflies, Fritz Miiller has discovered that there 

 exists a sexual allurement of a j)e(;uliur odour given out by 

 s])ef'ial patches of scales in individuals which are very 

 brilliantly coloured. This fact, he remarks, " is a clear indica- 



