220 
BULLETIN OF THE BUREAU OF FISHERIES 
All or nearly all are essentially shore fishes, at least in their breeding seasons. 
None of them, so far as known, spawns in deep water. Some spawn in the surf 
along the shore, some among the seaweed in quiet water, others in brackish water, 
and still others, like all members of the salmon family, ascend fresh-water streams 
for breeding. It is during the breeding season only that some of them are caught; 
and in the instance previously mentioned of the dependence of the cod fishery upon 
one of the species, it is when the little fish are approaching the shore to spawn. 
The genera comprised in the smelt family (Osmeridge), according to Jordan’s 
latest “Classification of Fishes” (1923), are principally the smelts (Osmerus), several 
species; capelin (Mallotus), 1 species; the eulachon (Thaleichthys) , 1 species; and 
the surf smelts (Hypomesus), 2 or more species. The latter two genera are peculiar 
to the Pacific. The others are common (generically) to the Pacific, Arctic, and 
Atlantic. This paper is concerned only with the genus Osmerus, principally with 
the smelts of the Atlantic, and more particularly with those of the Atlantic coast of 
North America. 
Opinions vary concerning the number of species of smelts comprised in the genus 
Osmerus. Whatever the number of species, they are all very closely related and not 
easily distinguished at a glance. Mary Fisk (1913) recognizes three species on 
the California coast — Osmerus thaleichthys, 0. attenuatus, and 0. starlcsi. The 
most northern of the Pacific smelts and the one most closely related to the Atlantic 
forms is technically known as Osmerus dentex. Besides these is the little known 0. 
albatrossis, Jordan and Gilbert, for which Jordan (1919) made the new genus Eper- 
lanoi. Smelts that have been identified as this species have been caught at Point 
Barrow, Alaska, at the mouth of the Mackenzie River, and on the northern coast of 
Siberia. A later classification (Hubbs, 1925) causes the genus Spirinchus (Jordan 
and Evermann, 1896, p. 522) to comprise 0. thaleichthys and 0. starlcsi, and he makes 
a new genus, Allosmerus, for 0. attenuatus. This arrangement leaves 0. dentex as 
the only species of Osmerus on the Pacific coast. 
There are, or have been, ichthyologists who regard the smelt of both sides of the 
Atlantic as specifically identical and identical with the above-mentioned 0. dentex. 
Futhermore, both in northern Europe and eastern North America there are smelts 
permanently resident in fresh-water lakes. As a rule, these also have been considered 
as specifically indistinguishable from the marine smelt. Thus Smitt (1895) does not 
distinguish between the smelts of certain lakes of Sweden and smelts of the Baltic, 
although he recognizes and shows that the Arctic-European smelt (Murman and 
White Seas) noticeably differs from the common smelt of Sweden. 
In this country smelts from two lakes in Maine were long ago described and 
named as specifically different from the marine smelt and also from a fresh-water 
smelt from another Maine lake. Years ago Foster and Atkins (1868) suggested 
that there might be many distinct species in the fresh-water lakes of Maine. 
Aiming at a solution of this question, hundreds of smelts have been studied by the 
present writer, and a great many data remain to be analyzed. This is reserved for 
a future paper. Herein the marine and fresh-water smelts are discussed separately 
but in a very general way. 
