2 H. C. FALL. 



The generic characters of Diplotaxis have been sufficiently 

 set forth in the books and need not here be given in detail. 

 Briefly stated, the more important ones are as follows: 



Ligula connate with the mentum; apical margin of pro thorax with 

 membranous border; ventral segments subconnate, but with distinct 

 sutures, five in number, the sixth not normally visible, the fifth connate 

 with the propygidium without suture; pygidium small, middle and hind 

 tibiae with two spurs, tarsi with distinct bisetose onychium and equal 

 claws, the latter variably toothed; anterior coxae conical and prom- 

 inent. 



In the LeConte and Horn classification the tribe Diplotaxini 

 comprises four genera, Orsonyx, Diazus, Diplotaxis and 

 Alobus. Of these Orsonyx is founded on a secondary sexual 

 character of the male and is therefore untenable; the female 

 is a normal Diplotaxis. The unique type of Alobus looks 

 greatly like Diplotaxis harperi, and is probably only an ab- 

 normal specimen of that species. Diazus is based on a single 

 rather small species with simple claws and 9-jointed antennae. 

 We have one species of Diplotaxis with 9-jointed antennae, but 

 none with simple claws. The genus may be considered valid 

 though not very securely founded. 



The tabulation of a genus of nearly one hundred species 

 is bound to present some difficulties, which are lessened in 

 proportion to the possibility of splitting up the mass into 

 sharply defined smaller groups. LeConte attempted to do 

 this in the synopsis above cited, depending chiefly on the 

 position of the ungual tooth, the prominence of the posterior 

 abdominal spiracle, and the form of the clypeus. His success 

 was only moderate, even with the relatively small number of 

 species treated. In the much larger series now involved I 

 have vainly attempted to separate the species into natural 

 groups, neither the characters used by LeConte nor any that 

 I can discern permitting its accomplishment. There seems 

 to be an entire lack of definite association of taxonomic char- 

 acters, which are so completely intercurrent that it is virtually 

 correct to say that if any one of these common to a considerable 

 number of species be selected as a point of departure, all of 

 the others will be represented in the group of species thus 

 segregated. Of the basic characters used by LeConte the 



