466 Mr, W. N. Lockington on American Alphei. 
But is it not the case with almost every genus except such 
as stand alone, forming of themselves a family or subfamily, 
that some of the less typical species intergrade with an adjoin- 
ing genus, or even with three or four adjoining genera? What 
is a genus? Is it not, like a species or family, a portion of 
the scale of Nature marked off arbitrarily for convenience in 
classification ? 
If there are rigid genera, not intergrading with others, it is 
because the connecting links have not yet been discovered—or 
because the causes which produced the generic characters were 
sudden and profound, destroying all races which did not change 
with sufficient swiftness. 
The multiplicity of genera with one species has been caused 
in great part by the too great subdivision which has been in- 
dulged in by naturalists. Characters merely specific have been 
made generic, while every slight variation of form or colour 
has added a species. More advanced modern naturalists, 
reviewing these co-called species with ample material, have 
proved that they are merely geographical varieties of the same 
species; and the result has been that we have numerous single 
species with a full binomial to themselves, yet with little to 
warrant such distinction. This is especially the case among 
the birds of North America; and will occur in other classes 
as soon as our knowledge is sufficiently advanced to detect 
the intergradation of allied forms from different localities. 
Were the test of geographical variation apphed to our 
North-American Alphei, I have little doubt that many would 
sink into varieties. Yet the forms actually distinct are really 
numerous, and the genus so large that it needs subdivision. 
Many of the species included in it, were they classified on the 
same principles with birds, would become genera. 
Previous to my notice of A. bellimanus, A. equidactylus, and 
B. longidactylus (in the Proc. Cal. Acad. vol. vii. part 1, 
pp. 34, 35) no species had been described from this coast ; 
immediately afterwards I added two other species ; Kingsley 
(Bull. U.S. Geol. and Geogr. Surv. vol. iv. no. 1, p. 189) 
adds six new species, all occurring at Panama, and notes the 
occurrence of the Atlantic A. heterochelis at Panama and 
Realejo, on the Pacific coast ; and in this paper I describe four 
additional species, noting also the occurrence of A. heterochelis 
on both shores of Lower California. 
IT am not aware that commensalism has previously been 
observed in this genus (Van Beneden, in his ‘ Animal Para- 
sites and Messmates,’ does not mention it), although Pon- 
tonia, an allied genus, is commensal; yet one at least of the 
Pacific species, B. equimanus, is a commensal under the mantle 
