1886 WORK ON THE GENTIANS 465 



evidence from a Chinee " Gif him four dozen, someting 

 vill transpire." At any rate the Chinee transpired, and 

 I hope some official will 



Here beginneth more gentian craze. 



I have not examined Moorcroft. yet, but if the figure 

 in Roxb. is trustworthy it's a primary and no mistake. 

 I can't understand your admitting Amarellae without 

 coronae. The presence of a corona is part of the defini- 

 tion of the amarella group, and an amarella without a 

 corona is a primary ipso facto. 



Taking the facts as I have got them in the rough, 

 and subject to minor verifications, the contrast between 

 the Andean, Himalayan, and Caucasian Gentian Florae 

 is very striking. 



Siraplices. Ciliatae. Coronatae. Interlobatae. 



Andes 27 (?) 15 2 



Himalayas 1 (Moorcroft.) 4 32 



Caucasus Pyrenees 2 (lutea 2 5 21 

 (all one) umbellata) 



I don't think Ciliatae worth anything as a division. I 

 took it as it stood. 



It is clear that migration helps nothing, as between 

 the old-world and S. American Florae. It is the case of 

 the Tapirs (Andean and Sino- Malayan) over again. 

 Relics of a tertiary Flora which once extended from S. 

 America to Eurasia through N. America (by the west, 

 probably). 



I see a book by Engler on the development of Floras 

 since tertiary epoch. Probably the beggar has the idea. 

 Ever yours, T. H. HUXLEY. 



GODALMING, Sept. 25, 1886. 



MY DEAR FOSTER We are here till to-morrow on a 

 visit to Leonard, seeing how the young folks keep house. 

 I brought the Egyptian report down with ma It is 

 VOL. II 2H 



