DEPARTMENT OF MARINE BIOLOGY. I35 



wide and sharply pointed and provided with a long protoplasmic rostrum, 

 while the tail is narrow and blunt. The cell-body is cylindrical and tapers at 

 either end, but its breadth is considerably increased by the presence of two 

 undulating membranes, which pass down from the rostrum on either side 

 and fuse at the posterior end. The interior of the cell is filled with cubical 

 bodies, which are larger in the middle and smaller at either end. These 

 bodies disappear either wholly or in part after the spermatozoa have been 

 in sea-water about two hours. They seem to go into solution and in their 

 place is left a brownish semi-fluid, granular substance in which may still be 

 seen the hexagonal outlines of the bodies. 



In sea-water the eupyrene spermatozoa soon come to rest, but the oligo- 

 pyrene continue active for about two hours. The latter are equipped with 

 neither cilia nor flagella, and their movement is caused solely by the undu- 

 lating membranes. At first large, slow waves pass down the membranes 

 alternately on either side ; later the M'aves become shorter and more rapid 

 and pass down both sides simultaneously. By this time, however, movement 

 of the spermatozoon itself has ceased, as it very soon attaches itself by means 

 of the rostrum. As soon as the cell flattens out a number of overlapping 

 folds are to be seen in the membranes ; these are due to the fact that the 

 membranes are not flat but wavy. The reaction of these membranes to salt 

 solutions isotonic with sea-water are uncertain ; it is not ciliary and it gives 

 some indication of being neuro-muscular. In NaCl movements of the mem- 

 branes may be observed at the end of five hours, while in MgClg they were 

 inhibited if anything. 



A subsequent study of the testis of Stromhus has yielded some interesting 

 facts. In the first place, the oligopyrene spermatocyte develops directly into 

 the spermatozoon. I have been able to find no indications of a division or 

 divisions such as occur in Paludina, nor have I seen any two spermatids 

 which could possibly be construed as daughter cells resulting from such a 

 spermatocytic division. I have not seen all the stages of the transformation, 

 but it appears to occur in this way : The nucleus of the large and easily recog- 

 nizable oligopyrene spermatocyte breaks up into a great number of vesicu- 

 lated fragments (chromosomes?), which come to lie scattered throughout 

 the cell-body. The cell then begins to elongate and the nuclear fragments to 

 degenerate. Later on a thick bundle of fibers is to be seen running through 

 part of the cell-body and extending beyond the cell-wall as flagella. These 

 probably arise from a large centrosome, which is visible before the nucleus 

 breaks up, but whose subsequent fate I have not as yet succeeded in tracing. 

 At all events, long before the spermatozoon has reached maturity, both the 

 bundle of fibers and the flagella have disappeared as such, fusing to form the 

 undulating membranes and the rostrum. The development of the cubical 

 bodies is very similar to that of the granules in the nurse-cells in Littorina 

 nebulosa. They begin to appear about the time of the disappearance of the 

 flagella and are entirely of a cytoplasmic origin. 



The evidence furnished by Strombus only goes to strengthen my belief 

 that the so-called oligopyrene spermatozoa are motile nurse-cells which are 

 quite comparable to those of the Littorince, if not entirely homologous with 

 them, but which only function as such in the seminal receptacle of the fe- 

 male. An examination of the seminal receptacle of Urosalpinx in midwinter 

 showed it to be full of both kinds of spermatozoa, indicating that the period 

 of copulation precedes that of oviposition by a considerable length of time. 

 Granting that in all cases one year's crop of spermatozoa fertilizes the next 



