1901.] FROM BltlTISH EAST Al^RICA. 349 



Lumbricidffi. The head- end is in several respects reiinDiscent of 

 the correspoiidiBg region of the sperinatophore of i»A./«- >■«'"- 

 Jormi. There is a terminal beak, which is toUowed by a thick 

 collar, after which the-at this part uarroxv-sperniatophore 

 .rradually widens out into the bag-like posterior region Ihe 

 ^permatophores are so bulky that there is only room tor two m 

 the larcve spermathecal sac ; at least I have not seen more than 

 two though the sac might accommodate perhaps three with some 

 little difficulty. In every case the mouth of the spermathecal sac 

 was plugged with one spermatophore, while another lay further up 

 the sac and not in contact with the first spermatophore. Lankester, 

 as I have already mentioned, found in TiMfe.v nvidonm th^t the 

 head of the spermatophore was moulded by the terminal part ot 

 the spermatheca ; he i)roved this not merely liy the correspondence 

 in form, but by the actual occurrence of a spermatophore with the 

 proiections of the head fitting into the lateral depressions ot the 

 spermatheca. I find precisely the same thing in these spermato- 

 phores of StMmannia. The wing-like processes ot the head fit into 

 concavities and on to convexities on the walls of the spermathecal 

 sac with great accuracy, while the beak-like anterior termination 

 of the sperinatophore corresponds to the narrow terminal duct ot 

 the sperm-receptacle. There can be, therefore, m my opinion, but 

 little doubt that this part of the spermatophore at least is moulded 

 bv the form of the spermathecal sac. The rest of the spermato- 

 phore also shows evidence of being moulded by the spermathecal sac. 

 The sac is narrov\'er at first and then widens out. In the same 

 way the spermatophore is narrower at first and afterwards becomes 

 broader. Its diameter is throughout not far short of the sac in 



which it lies. 



The only alternative locality for the formation ot the sperma- 

 tophore in StuUmannia is the spermiducal gland, or possibly 

 the unpaired muscular sac which opens on to the exterior m 

 relation to the spermiducal glands. But the lumen of neither 

 of these organs has anything hke the requisite breadth tor the in- 

 clusion of the f ullv formed spermatophore, which cannot therefore, 

 so far as I can see; be possibly moulded in its entirety anywhere save 

 in the spermathecal sac. This, however, is not tantamount to saying 

 that the spermatophore is altogether formed m the spermathecal 

 sac But before discussing the actual place of its origin, the rest 

 of its structure must be dealt A^-ith. The walls of the spermato- 

 phore are much thicker at the collar region than elsewhere : m 

 front of the collar the thin and narrow beak has thin walls, and is 

 widely open at the end. This, it must be remembered is the end 

 which is turned towards, and indeed is not far from, the external 

 orifice of the spermathecal sac. In examining closely with a 

 hieh power the end of the spermatophore, I observed a stream ot 

 spermatozoa which had evidently issued from the open mouth 

 Why this apparent waste takes place I do not know, but that it 

 must under normal circumstances take place is clear from the 

 width of the mouth, which is not narrow enough to keep the active 



