I 



EXPLANATION OF FORM AND COLOURING. 383 



some cases cumulatively brought about, or contributed to, a very high 

 degree of distinctiveness. This view was suggested to me in July, ]911, 

 by my actual experimental results up to that date — in other words, 

 by a consideration of the two lists which I append, and others like them. 

 That on roller B is reinforced, so far as some of the lower grades 

 go, by the unanimous verdict of animals on which I tested Acrseinse more 

 thoroughly than on that roller. I give the tests as I originally worked them out, 

 taking them from my letter of Feb. 16th, 1912, to Mr. Marshall, which he has 

 kindly lent to me. I have worked out the results of each experiment far more 

 exhaustively since and with a better knowledge of at any rate one complicating 

 factor. The work of bringing these results together into one list like the one 

 appended is not quite complete at the moment of my leaving England. I will add it 

 later if I am not in time to replace the present list by it. It will, I believe, entail 

 a few slight alterations in the first four grades. Otherwise (and in the main in 

 those grades) the present list is approximately correct. 



Preferences of Cobacias gabeultts. B. 



\ = not lower, possibly higher. 

 \ = ,, higher, ,, lowei'. 



Eioi'cdia (nmna and ivahlbergi) and 

 Atella, fir-st impression; U. neo- 

 2)hron, A. galene, Salamis spp. 3, 

 C. pillaaiia, Arrugia hasuta. 



1. Charaxes spp. (quite 12 spp. tried, in-1 



eluding neanthes and ethalion). 



2. Precis spp. 10, CVe?iis spp. 2, C.cloanthe, Y 



R. libeon, P. carchoi, H. dcedalus, M. 

 leda, Lachnoptera ayresiC^), Pyr. jywpfi.^ 



3. C.Jiorella, P. hippocoon, antheus^ and policenes, poss."^ 



ophidicephcdus (? ^^ )• I ^- hottentota (? "f ) and 



4. H. hippomene, H. perspicua, Hesperia spp. (? ^ ), R. f ^'.^^esi^s hereabouts. 



forestan and pisistratus, j 



5. P. lyceus and, if anything, slightly less liked, P. demodoctijS. 



6. P. dardanus S and Euxanthe loakefieldi. 



7. L. argia. Herpainia eriphia apparently hereabouts. 



8. L. thcdassina, P. angolanus, prob. P. leonidas, and, later placings, E. ivahl- 



bergi and ^nisipjnos. 



9. E. dry ope, P. eclierioides. 



10. E.liiarhas,P.natalensis (wet), and ^feZZ«, latest placings, hereabouts. 1 Cyrestis 



11. iV. agatha, N. sioynne^'toni, R. forestan with Amauris smell. J camillus. 



12. N. saclava, P. lucretia and trimeni; also, somewhat variably in 11, 12, and 13, 



Byblia spp., Teracolus spp., Phrissina isokani, and C. saha (latter poss. ^ ). 



13. N. goochi and, apparently, Alcena, Teriomi7na, Physcainura, and P.j)eucetia and 



tropicalis. 



14. M. campina, Ypthima nr. impura, Cycl. metis. M. sajitza prob. \ . 



15. Belenois spp. 4. Azaniis spp. (and poss. T. plinitcs) here or even, possibly, 



lower. 



