XII.] CONJUGATION AND SEXUAL REPRODUCTION. 1 63 



twins, can be exactly alike in this latter respect. Furthermore, 

 as a rule, no changes made their appearance in course of the 

 numerous generations during which the examination lasted, 

 with an exception which will be immediately described. I now 

 possess colonies of ^, as well as of 5, which cannot be dis- 

 tinguished from their ancestors in 1884, and which have there- 

 fore retained precisely the same markings as those of the 

 original animals. If we reckon six generations to the year, — 

 a number by no means excessive for breeding which took 

 place in a room, — about forty generations will have been passed 

 through since 1884. 



I attempted at first to produce the two forms by artificial 

 selection, breeding from the darkest individual of a colony of 

 the variety A, and from the Ughtest of a colony of B, in the hope 

 that, perhaps, in the course of generations, one variety might 

 be changed into the other. But I obtained no decisive results, 

 perhaps because I did not make my selection rightly ; for the 

 individuals are so very similar that it is often difficult and 

 indeed hardly possible to decide upon those which possess the 

 larger spots : perhaps also I mistook transient differences for 

 inherited ones, — a confusion which, naturally enough, cannot 

 be avoided. 



I was therefore all the more astonished to find, in 1887, 

 some individuals of the dark green variety B in the same 

 aquarium with the light variety A, and therefore side by side 

 with typical, light, clay-coloured individuals. At first I thought, 

 although it was most improbable, that these had been accident- 

 ally introduced, but the greatest care had always been exercised 

 in all these experiments. Furthermore, after the most pains- 

 taking precautions against such accidents, precautions which 

 prevented all possibility of the eggs being misplaced, there 

 presently appeared another similar case in a different aquarium 

 containing the variety A^ and, later on, yet another. In this 

 last case it was possible to find in the aquarium intermediate 

 forms between the two varieties, which had been wanting 

 on the previous occasions. Again, in May of the present year, 

 189T, another case was observed in which a single animal, dis- 

 tinctly belonging to the dark sub-species, suddenly appeared 

 among 540 mature Cyprides of the light variety. Five descen- 

 dants of this individual closely resembled their mother. 



M 2 



