XII.] CONJUGATION AND SEXUAL REPRODUCTION 21 J 



particular cases. In those Daphnids which, Hke the species of 

 Moina, inhabit small rapidly filled, but also rapidly drying 

 pools, the number of purely parthenogenetic generations which 

 succeed one another after the foundation of the colony, is 

 small. In Moina paradoxa and M. redirostris males appear 

 in the second generation, and some of the females produce 

 resting-eggs which require fertilization. If this did not occur, 

 if sexual reproduction, viz. multiplication associated with am- 

 phimixis, did not take place very soon after the foundation of 

 the colony, it would frequently happen that the latter would be 

 destroyed by sudden drought, without the formation of resting- 

 eggs to carry life in a latent condition over an unfavourable 

 period, and the colony would simply perish. It may be urged 

 that parthenogenetic eggs might have been provided with resting 

 shells like those which are, as a matter of fact, found in other 

 Phyllopods, for example Apus. But clearly the object is to 

 confer upon the species the advantage of periodically repeated 

 amphimixis, and this is therefore connected with the for- 

 mation of resting-eggs, and reproduction is so regulated that 

 the number of parthenogenetic generations is determined by the 

 average duration of the favourable periods of life. Thus, among 

 the marsh-dwelling Daphnids numerous purely parthenogenetic 

 generations succeed each other before a sexual generation 

 appears, while in those which inhabit lakes and are subject to 

 uniform conditions of life interrupted only by the cold of winter, 

 the cycle is still longer. In some species amphimixis may be en- 

 tirely abandoned, and this seems to occur most readily in those 

 which produce but one kind of egg, which must naturally be 

 provided with a protective resting shell, rather than in those 

 forming two kinds of eggs, of which only one is a resting-egg 

 and requires fertilization. Thus it is well known that most of the 

 colonies of the common Apus cancriformis are purely partheno- 

 genetic, and the same is true of the greater number of fresh- 

 water Ostracodes. 



Ten years ago, when I first directed my attention to the 

 parthenogenesis of these minute Crustacea \ I was able to dis- 

 tinguish three stages of reproduction,— the first was found in 



* Zoologischer Anzeiger, i88o, p. 72. ' Parthenogenese bei Ostra- 

 coden.' 



