﻿2 PROCEEDINGS OF THE NATIONAL MUSEUM vol. 102 



Gurney (1924, p. 559) says about the taxonomy of the Notostraca, 

 "The separation of species within the genera Apus and Lepidurus is a 

 matter of quite unusual difficulty, since there is much variation, and, 

 as a rule, the only characters on which separation can be based are 

 small and ill-defined." Barnard (1929, p. 230) agrees with him, and 

 so do I. 



As for other phyllopods, the females of the Anostraca present a 

 similar difficulty, as, sometimes, if they occur without males, it is 

 not even possible to say with certainty to what family they belong. 

 But the males of this order give excellent specific characters in the 

 genital organs and the second antennae — organs which display a 

 striking uniformity in the Notostraca. In the Conchostraca, the 

 structure of the head and the carapace provides rather good spe- 

 cific characters, but in these respects, too, the Notostraca are fairly 

 consistent. 



The characters actually employed — helpfully reviewed by Barnard 

 (1929, pp. 230-234) — for the most part represent differences of degree 

 only. A striking example of this is Packard's (1883, p. 319) key to 

 the American species of Apus; a study of the type specimens of the 

 species involved immediately reveals that the number of spines on 

 the supra-anal plate, which seems to be the fixed point in this key, 

 is not constant but overlaps the boundaries between his species. 

 According to my experience, otie seldom finds a specimen which is 

 in complete accordance with the description of any of these species, 

 and similar difficulties are met even when one goes to the type speci- 

 mens. These are not numerous enough to cover the range of indi- 

 vidual variation, nor are their descriptions extensive enough. Yet 

 such variation is especially important when the characters involve 

 dift'erences of degree. 



Of com'se, one can often compare material from neighboring locali- 

 ties in order to learn something of this variation, or, better still, one 

 can try to get abundant material from the original locality. The 

 first course is generally the only one practicable in museum work, and 

 I have tried it here. It is not very satisfactory, though some results 

 seem to be obtainable in this way. 



Each sample of Notostraca ought to contain about 50 or more spec- 

 imens in order to cover most of the variation. And, in my opinion, 

 a new species should never be described from less than 100 specimens 

 from the same district, preferably taken at varying times of the 

 season. Paratypes are, in many respects, just as important as 

 holotypes. 



A special question arises when we try to interpret the variation. 

 Ghighi (1924) suggests that the variation may sometimes be due to 

 the presence of two species in the same pool. I think that this occurs 



