360 BOTANY [Chap. XI 



Letter 683 degree. It is quite new to me what you show about the 

 effects of relationship in hybrids — that is to say, as far as 

 direct proof is concerned. I felt hardly any doubt on the 

 subject, from the fact of hybrids becoming more fertile when 

 grown in number in nursery gardens, 1 exactly the reverse of 

 what occurred with Gartner. The paper on Termites is even 

 still more interesting, and the analogy with cleistogene flowers 

 is wonderful. 2 The manner in which you refer to my chapter 

 on crossing is one of the most elegant compliments which I 

 have ever received. 



I have directed to be sent to you Belt's Nicaragua, which 

 seems to me the best Natural History book of travels ever 

 published. Pray look to what he says about the leaf-carrying 

 ant storing the leaves up in a minced state to generate 

 mycelium, on which he supposes that the larvae feed. Now, 

 could you open the stomachs of these ants and examine 

 the contents, so as to prove or disprove this remarkable 

 hypothesis? 3 



Letter 684 To F. Miiller. 



Down, May 9th, 1877. 



I have been particularly glad to receive your letter of 

 March 25th on Pontederia, for I am now printing a small 

 book on heterostyled plants, and on some allied subjects. 

 I feel sure you will not object to my giving a short account 

 of the flowers of the new species which you have sent me. 

 I am the more anxious to do so as a writer in the United 



valuable experiments on hybrid Abutilons, that the union of brothers 

 and sisters, parents and children, and of other near relations is highly 

 injurious to the fertility of the offspring." The Termite paper is in the 

 same volume (viz., VII.) of the Jenaische Zeitschr. 



1 When many hybrids are grown together the pollination by near 

 relatives is minimised. 



2 On the back of his copy of Midler's paper Darwin wrote : "There 

 exist imperfectly developed male and female Termites, with wings much 

 shorter than those of queen and king, which serve to continue the species 

 if a fully developed king and queen do not after swarming (which no 

 doubt is for an occasional cross) enter [the] nest. Curiously like cleisto- 

 gamic flowers." 



3 The hypothesis has been completely confirmed by the researches 

 of Moller, a nephew of F. Midler's : see his Brasilische Pilzblumcn 

 (Bo/an. Mittkeilgn. aus den Tropc?i, hrsg. von A. F. W. Schimper, 

 Heft 7). 



