ARTIFICIAL PARTHENOGENESIS IN ANNELIDS 675 



into sea-water whose osmotic pressure has been raised by the 

 addition of KC1 (e. g., 88 c.c. sea-water + 12 c.c. 2n KC1), 

 and if after two hours they are put back into normal sea- 

 water, they will develop into swimming Iarva3. In this case, 

 as a rule, only one embryo develops from an egg, and dwarf 

 larvae are an exception. If, however, instead of KC1 the 

 corresponding quantity of NaCl or MgCl 2 is added to the 

 sea- water, as a rule more than one embryo originates from 

 one egg, and larvaB of normal size are rare. I have not 

 made many experiments with CaCl 3 , but it seems to act 

 more like KC1 than like NaCl. In the experiments in which 

 the osmotic pressure of the sea-water was raised by cane- 

 sugar, dwarf blastulge were also observed. 



I have already mentioned in an earlier paper that the 

 lack of a membrane favors the origin of more than one 

 embryo from the unfertilized egg. The fertilized egg has a 

 membrane which keeps the cleavage cells together. But if 

 the membrane be destroyed, the egg may give rise to more 

 than one embryo. In a small number of unfertilized eggs 

 the treatment with KC1 gives rise to a very thin film, which 

 may act as a membrane and prevent the cleavage cells from 

 becoming separated. But such a fine film is lacking in the 

 majority of eggs treated with KC1 (or CaCl 2 ) in the right 

 proportions to produce parthenogenesis. And yet we do not 

 notice the falling apart of cleavage cells which in the case of 

 the NaCl eggs orMgC! 2 eggs leads to the formation of more 

 than one embryo from an unfertilized sea-urchin's eggs. 

 The observation of the process of cleavage shows that the 

 treatment of the eggs with KC1 increases their power of 

 adhesion. The various cleavage cells of a K egg stick to- 

 gether, while after a treatment with NaCl the cleavage cells 

 adhere much less to one another and fall apart. The same 

 tendency is produced by the addition of MgCl g to sea-water. 

 It is quite possible that the relative amount of the various 



