76 Journal New York Entomological Society. [Vol. vm. 



jointed, with a loose club which is generally 3-jointed, but sometimes 

 purely 2 -jointed, and, in one case — Anchor his — 4-jointed, a character 

 remindful of Tritomidas. The anterior coxal cavities are generally 

 widely open behind, but are completely and rather broadly closed in 

 JDip/oca'/us, completely but less broadly in Cryptophilus, narrowly but 

 almost completely in Hap/olophi/s, and about half closed in Setaria, 

 proving that no useful generalization in the definition of the family 

 can be drawn from the form of the cavities. The Biphyllini, as stated 

 by Reitter, are evidently a perfectly natural part of the present family, 

 this being proved by general organization, tarsal structure and espe- 

 cially by the radiating straight lines of the first ventral segment, also 

 occurring in Cryptophilus, and, in an arcuate form, in Tomarus* 

 The tribe is quite out of place in the Tritomidae, to which it was 

 assigned by LeConte and Horn. 



The Cryptophagidae comprise two distinct subfamilies as shown by 

 the following characters : — f 



Antenna; inserted under the acutely margined sides of the front and remotely sepa- 

 rated at base, the palpi mutually dissimilar, the maxillary elongate and slender, 

 with the fourth joint elongate and more or less acuminate toward the tip, the 

 labial short, with the last joint enlarged, oval to securiform ; pronotum generally 

 bifoveolate at base Cryptophagin^ 



Antennas inserted on the front and more or less approximate at base, the palpi mutu- 

 ally similar, short, stout and acuminate, the last joint of both small, narrow and 

 subuliform ; trochanters always narrow and elongate, bearing the femora dis- 

 tally ; pronotum never bifoveolate at base, though generally impressed ; anterior 

 coxal cavities always widely open behind, the tarsi invariably simple and fili- 

 form ; eyes always basal, the first abdominal segment never having radiating 

 lines Atomariin.b 



In tarsal structure these two subfamilies are linked together by way 

 of the Cryptophagini and Casnoscelini. The insertion of the antennas 

 in Anthcrophagus seems to suggest also a slight drift toward the Atoma- 

 riinae, but this is very feeble and more apparent than real. In the 

 mode of antennal insertion, and especially in palpal structure, the two 

 subfamilies are radically distinct ; perhaps species may be discovered 

 showing intermediate characters, but it is more probable that these 

 bonds have long ago become extinct. 



* These lines also occur in the subfamily Silvaninae of the Cucujidas. 

 f Names to which an asterisk is affixed apply to tribes or genera which do not occur 

 within the limits of the American fauna as far as discovered. 



