GERM CELLS IN PEDICELLINA AMERICANA 7 



That this view is probably expressive of the truth is to be 

 inferred from the existence, in the same and in alHed genera, of 

 all the necessary transitions between the primitive conditions 

 observed in the majority of forms, on the one hand, and the 

 specialized condition of P. americana and Ascopodaria macropus 

 on the other. Beginning with P. glabra, where the individuals, 

 as well as the colony, are all monoecious, the series leads di- 

 rectly to those where, as in P. ecJiinata proterandry or proter- 

 ogyny sets in, and finally becomes distinctive of the species. 

 In the genus Loxosoma, this is the all prevailing form in which 

 the hermaphroditism is found. Thus Harmer writes of this 

 genus : "I have invariably found that mature ovaries and testes 

 are mutually exclusive. It is easily shown that individuals 

 containing developing embryos in their vestibule are not pro- 

 vided with testes in the species oi Loxosoma and Pcdiccllina which 

 I have examined. In some cases, a vesicula seminalis contain- 

 ing spermatozoa is found, although the testes seem to be com- 

 pletely absent. This fact, perhaps, indicates that the male 

 gonads, which must have been originally present, have atro- 

 phied in order to make room for the development of the ova- 

 ries." We have now only to suppose that this proterandry or 

 proterogyny of the colony is distributed over several individuals 

 instead of being localized in one, and the condition found in 

 P. americana is obtained. The extension of the period between 

 the development of the germ-cells of the two sexes is probably 

 in response to some change in the relation of the zooids to the 

 colony as a whole in which process the latter becomes more 

 highly individualized. 



This condition can finally be traced one step further where, 

 as in a form like P, bclgica or Loxosojiia annelidicola, one or the 

 other of the sexes never develops, leaving both the polyps and 

 the colony dioecious. This is the end of the series. It has 

 therefore been shown that in the Endoproctous Bryozoa, all 

 possible relations of the germ cells in single individuals and 

 groups of individuals exist, and that from the primitive condi- 

 tion of complete monceciousness, a full series of transitions can 

 be traced to as complete diceciousness. There is, however, an 



