1885.] Geology and Paleontology. 591 
and the severity of the climate the only obstacles, we might ex- 
pect that many more of the forms which crowd the coasts would 
work their way up the rivers. As a rule, however, the fresh- 
water forms are sfa distinct from the marine, retain their dis- 
tinctness everywhere, and, in time, are well marked as far back as 
the Mesozoic. It is therefore probable that the fact that the 
majority of marine invertebrata are diffused by means of free- 
swimming larvæ has been one of the chief obstacles to their 
spread up the rivers. These fragile and feeble larvæ always swim 
along with even an ocean current, and are utterly powerless to 
stem that of a river. Should a siow-moving marine animal suc- 
ceed in ascending some distance up a stream, its larvæ, if free- 
moving, would infallibly be carried out to sea. By a detailed 
examination of the forms which inhabit fresh water, Professor 
Sollas shows that in most of them the free larval stages are 
suppressed. Other causes may exist. Thus the absence of suit- 
able food is sufficient to account for the lack of carnivorous gas- 
tropods and cophaliper: i in the rivers. 
Fresh-water animals may be converted into marine in three 
ways: (1) by direct migration; (2) by the conversion of the area 
they inhabit into a fresh-water basin or lake; (3) by adaptation to 
a terrestrial or marsh-loving habitat, and subsequent exchange 
of this for a fluviatile or lacustrine one. The first method can 
scarcely occur with fixed forms, unless they are parasitic upon 
locomotive animals. Some prawns and crabs appear to have thus 
immigrated by compliance with the three conditions, but the in- 
stances are very few. The wide changes in the distribution of 
land and water that have perir place in the course of geological 
time offer a more probable mode of the gradual transformation of 
a fauna from a marine to a e antes one. The comparative 
poverty of the latter may be due to the escape of some species, as 
well as to the extinction of others. The earliest lakes known are 
of the Devonian period, and one Devonian fossil at least, Azo- 
donta jukesu, has been found. Helicidæ are found in the coal 
measures, and are probably the ancestors of the Limnæidæ. In 
the Lias and Oodlite numerous fresh-water mollusks occur, and 
Cyrena, Neritina and Hydrobia probably date from the Trias. 
Saectal genera of fresh-water mollusks were already distributed 
over parts of the Palæarctic, Nearctic and Oriental regions in 
Cretaceous times. The Tertiary lakes of the northern hemi- 
sphere have suffered from a glacial era, and the Caspian has be- 
come unwholesome by concentration of its waters, yet it retains a 
relic of a Tertiary fauna; while the Central African lakes have a 
remarkable assemblage of Mollusca. 
No marine mollusk is known to pass through a “ glochidium ” 
stage, like that of the Unionide; no marine Polyzoon or sponge 
produces statoblasts; no marine Phyllopod an ephippium ; and 
no Tubularian an egg in a horny shell like that of Hydra. 
