1886 WORK ON THE GENTIANS 463 



I am inclined to think that I had much better be out 

 of it next year. The attempt to look over examination 

 papers now would reduce the little brains I have left to 

 mere pulp and, on the other hand, if there is any row 

 about results, it is not desirable that I should have to 

 say that I have not seen the answers. 



When I go you will probably get seven devils worse 

 than the first but that is not the fault of the first devil. 



I am picking up here wonderfully in spite of the bad 

 weather. It rained hard yesterday and blew ditto 

 to-day it is blowing dittoes but there is sunshine 

 between the rain and squalls. 



I hope you are better off. What an outlandish name 

 " Tetronila." I don't believe you have spelt it right. 

 With best regards to Mrs. Donnelly and my godson 

 Ever yours, T. H. HUXLEY. 



4 MARLBOROUGH PLACE, 

 Sept. 16, 1886. 



MY DEAR HOOKER I have sucked Grisebach's brains, 

 looked up Flora B. Americana, and F. Antarctica and 

 New Zealand, and picked about in other quarters. I 

 found I knew as much as Grisebach had to tell me (and 

 more) about lutea, purpureo-punctata, acaulis, campestris, 

 and the verna lot, which are all I got hold of at Arolla. 

 But he is very good in all but classification, which is 

 logically "without form and void, and darkness on the 

 face of it." 



I shall have to verify lots of statements about gentians 

 I have not seen, but at present the general results are 

 very curious and interesting. The species fall into four 

 groups, one primary least differentiated three, specialised. 



1. Lobes of corolla fringed. 2. Coronate. 3. Inter- 

 lobate (i.e. not the " plica " between the proper petals). 



Now the interesting point is that the Antarctic species 

 are all primary and so are the great majority of the 

 Andean forms. Lutea is the only old-world primary, 



