174 



SCIENCE. 



[N. S. Vol. XI. No. 266. 



are established when the egg has only forty- 

 eight cells. 'I'he rosette cells are very large ; 

 and the ciliated prototroch, which consists 

 of sixteen cells, derived from the first set 

 of ' micromeres,' forms a complete girdle 

 around the egg. 



A typical trochophore is formed, of which 

 the plane of bilateral sj'mmetrj'^ corresponds 

 to a vertical plane bisecting 5 and D of the 

 four-celled stage. A postoral circlet of 

 strong cilia appear at a short interval be- 

 hind the prototroch, and a long tuft of 

 flagella is still earlier developed upon the 

 apical plate. There is no true paratroch. 

 Eye spots are present, and trochophores 

 and larvae are positively phototactic. 



No traces of metameric segmentation 

 manifest themselves throughout the course 

 of development, which was observed con- 

 tinuously until the young worms had 

 reached the age of seven weeks. This and 

 certain other embryological facts seem to 

 indicate that the Gephyrea are somewhat 

 more closely related to the Flatyhelminthes 

 than to the Annelida. 



Notes on the structure of Alma nilotica, a gilled 

 earthworm from Egypt : P. M. Eea. 

 Alma nilotica Grube, has been known 

 since 1855, but has never been thoroughly 

 investigated. Its systematic position is 

 uncertain, but the present research shows 

 conclusively that it is an Oligochaete hav- 

 ing many of the characteristics of the Geo- 

 scolicidae. The possibility of the identity 

 of this form with the genus Siphonogaster 

 of Levinsen increases the interest of this 

 remarkable worm. The material available 

 at present is sexually immature, but it is 

 hoped that specimens collected in the spring 

 will determine this point. A pair of ovaries 

 has been demonstrated in segment 13 and 

 testes in 10 and 11, but no evidence of the 

 enormous penial processes of Siphonogaster. 

 The gills, which are the most character- 

 istic feature of the worm, are outpocketings 



of the body wall, taking with them the 

 layer of circular muscles but leaving the 

 longitudinal muscles behind. They are 

 provided with afferent and efferent blood 

 vessels. The epithelium of the gills and 

 whole body surface is highly vascular. 

 The dorsal blood vessel extends no farther 

 forward than the seventh segment, where 

 it ends abruptly in the most anterior pair 

 of hearts. There is a supra- oesophageal 

 vessel and two remarkable lateral vessels 

 which will be more fully discussed in a 

 later paper. Connected with the lateral 

 vessels are numerous spherical acini, closely 

 approximated to the inner surface of the 

 body-wall, which appear to be identical 

 with the structures figured by Perrier as 

 occurring on the walls of the oesophagus in 

 Perichaeta. 



On the life history of Autolytus cornutus and 



alternate generation in annelids : P. C. 



Mensch. 



The claim for alternate generation in 

 annelids arises from investigation on the 

 Syllidians, chiefly Autolytus. It was first 

 suggested by Quatrefages and Krohn, but 

 for the first time fully described by A. 

 Agassiz for Autolytus cornutus. Agassiz re- 

 garded the parent stock as distinctly asexual 

 and in this manner described a true alter- 

 nate generation — the asexual parent stock 

 alternating with sexual stolons. 



The asexual condition of the parent stock 

 is, however, not constant and the percentage 

 of parent stocks with sexual products is 

 sufficiently great to strongly indicate that 

 the presence of reproductive products to- 

 ward the close of the phenomenon of bud- 

 ding is a constant stage in the life -history 

 of this Syllid. This being the case there 

 would be, not an alternation of generation 

 but at most only a sexual dimorphism. 



Another aspect of this question is pre- 

 sented by the morphological characters of 

 the stolon itself, in that the stolon does not 



