86 JOURNAL OF CONCHOLOGY, VOL. II, NO. 3, JULY, I9O4. 



on with them the following year, commencing earlier, but the untimely 

 death of Prof. Marshall put an end to the project, and I have not 

 since had an opportunity of going further into the matter — systematic- 

 ally at any rate. 



Development by means of pelagic larvae provided with a velum is 

 generally characteristic of marine Pelecypoda, but not of freshwater 

 forms. SphcBriiiin, Fisidiuju, and kindred genera develop their eggs 

 in brood pouches within the gills of the parent, until the formation of 

 the perfect animal is almost complete, and the young have already 

 assumed the adult form on leaving the parent. The ova of the 

 Unionidge also develop in the gills, and when the larvse are expelled 

 therefrom they have become so far developed, and furnished with 

 shells and shell-hooks, as to enable them to attach themselves to the 

 fishes upon which they live parasitically for a time — but they possess 

 nothing in common with the free-swimming larvae of Osfrea, Mytilus, 

 and other marine forms. 



The ova of Dreissensia are deposited about the middle of June. 

 They are remarkably small, with a scanty supply of yolk, and have an 

 extremely delicate envelope. The mussel during oviposition slightly 

 opens the valves of its shell and by snapping them quickly together 

 ejects a tiny ball of ova. This process is repeated at intervals and 

 may readily be observed in the aquarium. The balls of ova, which 

 look like little whitish lumps of mucus, may be seen lying near the 

 parents, and retain their cohesion for some time. The subsequent 

 developmental stages, and especially the unequal segmentation of the 

 ova, are processes which approximate very closely to the conditions 

 which obtain among the marine mussels. 



As the ova gradually develop the embryo first assumes a roundish 

 oval shape, and then the shell-gland, an organ of extreme importance 

 in the Pelecypoda, appears, and the embryo — or more correctly 

 speaking the larva — shews as a broadly pear-shaped body, almost 

 indistinguishable from the marine larvae of Modiolaria or Cardium. 

 This is the "Trochophore" stage, which in Sphceriuvi, Pisidmm, and 

 Gastropods is passed through in a more or less modified form within 

 the egg-membrane, but is preserved as a free-swimming or veliger stage 

 in the case of Dreissensia. Until their discovery by Dr. Korschelt 

 veliger larvae were quite unknown in fresh water. 



The development of the larva goes on until the shell-membrane has 

 considerably increased in size, and presents the appearance of the 

 bivalve mussel shell (fig. i, s.). This condition continues for a long 

 period, during which the characteristic velum, a fleshy organ fringed with 

 cilia at its margin, appears. By means of the expanded velum, and 

 ciliary movement, the larva can move very rapidly; and often when 



