{ cxi ) 



from races by preferential syngamy is of such high import- 

 ance that we may confidently hope that the attention here 

 directed to the question, and especially the quotation of 

 Darwin's letters to Bates, may lead to that " exact and 

 prolonged observation," accompanied by careful records, with- 

 out which a safe decision cannot be reached. In the 

 meantime tlie decided impressions of two such naturalists as 

 H. W. Bates in South America and Roland Trimen in South 

 Africa render it in every way probable that the conclusion 

 will be established on a firm foundation.* 



It is also possible that asyngamy may be brought about by 

 the breaking of what we may call " a syngamic chain." In the 

 case of large and widely-distributed interbreeding communities, 

 it is an open question whether syngamy would freely take 

 place between the most distant of the outlying sections if 

 directly brought into contact, and whether, even if syngamy 

 prevailed, there would be any diminution in fertility. 



Limnas chrys%p%>us, perhaps the commonest butterfly in the 

 world, forms a probably continuous syngamic chain stretching 

 from the Cape of Good Hope at least as far as Southern China. 

 It is even reported from Japan. The far Eastern forms are 

 readily distinguishable by the greater size of a single white 

 spot, giving quite a different appearance to the fore-wing. 

 If pupae or eggs were transferred from Hong-Kong or Macao 

 to South Africa, would the perfect butterflies freely interbreed 



* Dr. T. A. Chapman sends me the following interesting and suggestive 

 note : — 



"I met lately with a curious instance that deserves following up, of 

 some bearing on the question of selective mating of varieties. 



"I saw some broods of P. pldxas lately that differed from each other, 

 but each brood was remarkably uniform. There were three broods, all 

 bred in the same conditions, in a gieenhouse (by Mr. Carpenter of 

 Leatherhead). It seems difficult to explain this, unless both parents of 

 each brood were very nearly identical. 



"Mr. Fi'oliawk, who has bred the species largely, tells me he has 

 noticed similar facts. 



"When I bred Aa-onyda tridens and jisi largely, some fifteen or more 

 years ago, I noticed that each brood had its own pairs, and suggested that 

 tridens was now trying to break up into separate species just as some 

 ancestor split into psi, tridens and cuspis. 



"Another fact I observed in Acronycta rather bears on the other side 

 of the question. Of A. strigosa I reared a large brood, which paired 

 readily and frequently together, but no eggs were laid. I then got some 

 captured males, which paired with equal readiness with the bred females, 

 and as a result obtained plenty of fertile eggs." 



