j8 CHILD. [VOL. I. 



The larva of Arenicola possesses a paratroch (Fig. 12, Par), 

 but it is not functional until a very late stage. 



Dr. Mead has very kindly furnished me with the lineage of 

 the paratroch of Amphitrite, and it is interesting to note that 

 the origin of this organ in the two cases is entirely different. 

 The paratroch of Arenicola arises from certain derivatives of 

 X, but not only is the series of divisions different in direction 

 from that found in Amphitrite, but the cells, or a part of them, 

 pass through several more generations than in Amphitrite. 



The free-swimming trochopore of Arenicola already possesses 

 three trunk segments with setae. The pelagic life does not 

 last over two days. The larva 1 sinks to the bottom and crawls 

 slowly about, and the growth of new segments begins. Fig. 12 

 is an outline of the trochopore in lateral view, the segmenta- 

 tion not being shown. The object of the figure is to show the 

 area of the ectoderm formed by X. 



Just a word here in regard to the cleavage of Sternaspis 

 scutata. My work on this worm was begun at Naples, but, 

 owing to loss of material, I have never yet been able to com- 

 plete it. I desire to express my sincerest thanks to Professor 

 Agassiz for the privilege of the Agassiz table at Naples, and 

 also to Professors Dohrn and Eisig, as well as the other mem- 

 bers of the staff of the Zoological Station, for their kindness 

 during my stay. The unsegmented egg of Sternaspis is almost 

 completely filled with very large yolk-spheres, but none of them 

 pass into the ectomeres or the mesoblast. All these cells are 

 relatively small, while the entomeres remain as enormous, yolk- 

 packed cells. Yet up to a stage of about eighty cells (as far as I 

 have been able to follow the cleavage of Sternaspis} the succes- 

 sion of cleavages is cell for cell the same as that of Arenicola. 

 Sixteen cells are formed, corresponding to the primary trocho- 

 blasts of Arenicola, but Sternaspis has no prototroch, and they 

 simply form a part of the ectoderm. On the other hand, the 

 larva never really resembles a trochopore. It is not free- 

 swimming. Vejdovsky ('81) states that the larvae at Trieste 



1 The larvae are very easily kept alive in covered glass dishes with a supply of 

 ulva. I have kept them in the laboratory for three months, and thus have been 

 enabled to preserve a complete series of stages of the late larval period and the 

 metamorphosis. 



