934 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxvi. 



No Astarte has radial ribbing, but man^^ develop in the adult state 

 crcnulations on the inner margin of the valves. Some species are 

 crenulate only when fully adult, others develop crenation at resting 

 stages, others are always without them. The teetli of the hinge are 

 frequently reversed as regards the valves. The laterals vary extremely 

 in the fullness of their development, and the outline of the valves is 

 also often ver}^ variable, most of the trigonal species having also 

 elongated or oblique varieties. The conditions in the boreal seas, 

 where these animals chiefly abound, seem to make for profusion in 

 individuals and paucity of species, a state of things obviously favor- 

 able to individual variation. 



The dullness of color characterizing the shells of this group is to 

 some extent made up for by the l)right colors of the soft parts which 

 are usually yellow, orange, or vermilion. The anal siphon is short, 

 complete, plain-edged and valvular. The branchial siphon is formed 

 l)v apposition of the ciliate border of the free mantle edges. The 

 foot is subquadrate, the gills small , reticulate, and free. The eggs are 

 ripe in April in the latitude of Long Island Sound, and are dis- 

 charged into the water for fertilization. The animals live partly 

 immersed in mud or sand and form a large part of the food supply of 

 the walrus and many fishes. The variabilit}^ previously alluded to 

 has made the identification of species difficult and their synonymy 

 almost hopeless. 



Dr. Jeffreys, who pul)lished much about the northern fauna, unfor- 

 tunately was disposed to lump together rather than discriminate, not 

 having the large series necessary for elucidating our American spe- 

 cies. Sowerby's monographs of this group are very unsatisfactory and 

 imperfect, covering less than half the known forms. The most suc- 

 cessful attempt at a review of the species is that of Mr. Edgar A. 

 Smith, of the British Museum, in 1881, in the Journal of Conchology. 



I have found in reviewing our American species that a more narrow 

 specific limitation removes some of the difficulties surrounding the 

 subject. While assenting to many of the views heretofore expressed 

 by naturalists, and disavowing any claim of finality for the decisions 

 arrived at, I have endeavored to discriminate the rc^cognizable Amer- 

 ican forms whether these be regarded as species or not. By adopting 

 names for them we at least have the satisfaction of knowing what we 

 mean when we employ a name, which is impossible under the system, 

 or want of system, of Dr. Jeffreys. For the same reason 1 have been 

 unable to avail myself of much work, systematic and distributional, 

 which is in print, since it is impossible to know which of several forms 

 is intended in a given case where a name covering a number of types 

 has been usQd. The distrilmtion mentioned in my list herewith is 

 taken from actual specimens and only exceptionally from the litera- 

 ture. The collection of Astartes from the boreal regions of the New 



