i88i.] PROFESSOR AGASSIZ. 421 



p. 241) on the lower and softest side of the body — and that 

 to sting a certain segment was found by far the most suc- 

 cessful method ; and was inherited like the tendency of a 

 bulldog to pin the nose of a bull, or of a ferret to bite the 

 cerebellum. It would not be a very great step in advance to 

 prick the ganglion of its prey only slightly, and thus to give 

 its larvae fresh meat instead of old dried meat. Though 

 Fabre insists so strongly on the unvarying character of in- 

 stinct, yet it is shown that there is some variability, as at p. 

 176, 177. 



I fear that I shall have utterly wearied you with my scrib- 

 bling and bad handwriting. 



My dear Romanes, yours, very sincerely, 



Ch. Darwin. 



Postscript of a Letter to Professor A. Agassi z^ May ^th^ 



1881 :— 



I read with much interest your address before the Ameri- 

 can Association. However true your remarks on the gene- 

 alogies of the several groups may be, I hope and believe that 

 you have over-estimated the difficulties to be encountered in 

 the future : — A few days after reading your address, I inter- 

 preted to myself your remarks on one point (I hope in some 

 degree correctly) in the following fashion : — 



Any character of an ancient, generalised, or intermediate 

 form may, and often does,. re-appear in its descendants, after 

 countless generations, and this explains the extraordinarily 

 complicated affinities of existing groups. This idea seems 

 to me to throw a flood of light on the lines, sometimes used 

 to represent affinities, which radiate in all directions, often to 

 very distant sub-groups, — a difficulty which has haunted me 

 for half a century. A strong case could be made out in 

 favour of believing in such reversion after immense intervals 

 of time. I wish the idea had been put into my head in old 

 days, for I shall never again write on difficult subjects, as I 

 have seen too many cases of old men becoming feeble in 



