212 BENJAMIN H. GRAVE. 



The rate of development and age at metamorphosis are shown 

 in Table I. The later stages were obtained by feeding the 



TABLE I. 



SCHEDULE OF NORMAL DEVELOPMENT AT MEDIUM SUMMER TEMPERATURES, 



20-2I C. 



First polar body 12 to 15 minutes 



Second polar body 28 to 35 



First cleavage 60 to 70 



Second cleavage 90 to 102 



Blastula 5 to 6 hours 



Gastrula 7 to 10 



Trochophore 12 to 18 " 



Metamorphosing trochophore 20 to 30 " 



Veliger complete 2 to 24 days 



Metamorphosis of veliger into adult form 16 to 24 " 



Sexually mature in one year 

 Fully grown in four years 

 Duration of life two to four years 



larvae, by means of which they were kept alive for a month. 

 One can distinguish the new parts of the shell, as added to the 

 original larval shell, and so tell where new growth has occurred. 

 It is quite likely that the development of the larvae in the open 

 ocean is more rapid and uniform than in the laboratory. How- 

 ever, the veligers seemed to thrive upon diatoms in a large 

 aquarium jar kept agitated by a mechanical device, and in 

 time a considerable number went through metamorphosis. 



The measurements given in Table III. show that the veligers 

 grew somewhat before metamorphosis but more rapidly after- 

 wards. This is not fully in line with data on the rate of growth 

 of Chiton (unpublished), which show that there is no measurable 

 growth until after metamorphosis. Chiton seems to be unusual 

 in this respect. The larvae of Cumingia did not develop uni- 

 formly in this aquarium habitat. Some grew symmetrically, 

 adding equally to the shell all around, and others grew asym- 

 metrically. I cannot, therefore, be sure of the normal form. 

 Moreover, there is very considerable variation in the proportions 

 of the normal veligers of a single female. If variation is as 

 extensive in all lamellibranchs as in Cumingia it is difficult to 

 see how the early larvae of different species can ever be distin- 

 guished satisfactorily by measurements, although Stafford has 



