1895.] The Genera of Branchiostomidae. 457 
THE GENERA OF BRANCHIOSTOMIDAE. 
By THEODORE GILL. 
The work of Mr. Arthur Willey on “Amphioxus and the 
Ancestry of the Vertebrates ” (N. Y., 1894) is a useful compila- 
tion of what is known respecting the general anatomical char- 
acteristics of the Branchiostomids, but much remains to be yet 
made known regarding structural details and the range of 
variation, not only within the family, but also within specific 
limits. A first step toward the proper examination of such 
variations is to segregate the species into groups distinguished 
by positive structural peculiarities or associations of charac- 
ters. For the expression of such structural peculiarities, generic 
diagnoses and terms are the best expedients, and they will 
differentiate most clearly characters of secondary importance 
from those of tertiary rank and the common or family charac- 
ters or those of primary rank.- Unfortunately, Mr. Willey has 
not distinguished between the various grades of characters, but 
has thrown all the representatives of the family into one genus 
without any sectional subdivision and (adapting the sequence 
of Dr. E. A. Andrews), has interposed “ B. cultellum” between 
“ B. caribaum” and “ B. bassanum,” and even (unlike Dr. 
Andrews) added to the genus the Asymmetron lucayanum of 
Andrews. There appear to me, however, to be at least three 
well-marked genera. These are Branchiostoma, Epigonichthys 
and A tron. Another (Paramphiozus) has been proposed by 
Prof. Haeckel (1898) for the “ Branchiostoma bassanum” of 
Günther, and it is gratifying to find that my views seem to be 
in accord with that eminent master of discrimination and 
valuation of morphological characters and their expression in 
diagnostic form. Doubtless Prof. Haeckel has good reasons for 
the genus Paramphioxus, but he has not yet formulated its 
characters, although he has indicated that it has unilateral 
gonads, and, such being the case, it must be related to Epi- 
gonichthys, although apparently distinguished from it by differ- 
ence in the relative development of the fins. A fifth genus is 
31 
