[445] INVERTEBRATE ANIMALS OF VINEYARD SOUND, ETC. 151 
Fuller. It may, nevertheless, occur annually in winter, and yet be sel- 
dom observed; for very few naturalists go out to collect marine animals 
in winter and early spring. 
The bivalve shells mostly produce minute young, or larvw, which are 
at first provided with vibrating cilia and swim free for several days, as 
is well known to be the case with the oysters, clams, muscles, Teredo, &c. 
But a few species, like the Tottenia gemma, (p. 359,) produce well devel- 
oped young, furnished at birth with a well formed shell. 
The common fixed Ascidians, both simple and compound, mostly pro- 
duce eggs that hatch into tadpole-shaped young, which swim about for 
a short time by the undalatory motions of the tail, but finally become 
fixed by the head-end, and losing, or rather absorbing, the tail-portion, 
rapidly develop into the ordinary forms of the ascidians. This pro- 
cess, although often very rapid, is a very interesting and complicated one 
In Molgula Manhattensis there is, according to the observations of Dr. 
Theodore A. Tellkampf, an alternation of generations. He states that 
the minute yellow ova were discharged July 18, invested in a viscid 
yellowish substance, which become attached to the exterior of many 
specimens. In a few days the “ viscid substance” had changed its ap_ 
pearance and became contractile; the ova became larger, round, and of 
different sizes; ‘“‘after two or three days the largest protruded some- 
what above the surface of the common envelope, and presented a circular 
or oval aggregation, like that of the Wammaria found a year ago;” on 
the 11th day, the round ova had increased in size, with a central round 
or oval orifice through which the motion of the cili of the branchial 
meshes were visible. ‘ The orifice had approached on the 1st of August 
more or less to one apex; in some specimens, which were now oval, it 
was terminal.” In this stage he names it Jlammaria Manhattensis, 
regarding the Mammaria as a “nurse ;” within each of the Mammarir, at 
the end opposite the branchial orifice, there was seen a mass of cells, 
which ultimately developed into a tadpole-shaped larva, similar to that of 
other ascidians. He observes that the Mammarie increase after the 
discharge of the larve, and that gemmation takes place within the 
common envelope.* These observations, if correct, are very interesting 
and important, but they need farther confirmation. The development 
of the larvee from the Mammaric into Molgula was not traced; neither 
did he witness the actual discharge of the ova, which produced the 
Mammaric, from the Molgula. They may possibly have no relation with 
one another. 
Several kinds of Ascidians, however, swim free in the water during 
their entire life. The most common Ascidian of this kind is the Salpa 
Cabotti, (Plate XX XIII, figs. 254, 255.) This, like the other species, 
exists under two different forms ; or, in other words, it is one of those an- 
imals having alternations of generations. The sexual individuals (fig. 
255) are united together into long chains by processes (c) from the sides 
* Annals of the Lyceum of Natural History of New York, Vol. 10, p. 83, 1872. 
