THE ANNALS 
MAGAZINE OF NATURAL HISTORY. 
[EIGHTH SERLES.] 
No. 106. OCTOBER 1916. 
XXX VII.—The Scales of the Brotulid Fishes. 
By T. D. A. Cockrretu, University of Colorado. 
In the ‘ Proceedings of the Biological Society of Washington,’ 
vol. xxvi. (1913) p. 76, I gave a brief account of the scales 
of two species of Brotulidz. Since that time, thanks to the 
kindness of Dr. Hugh M. Smith, I have been able to study 
the scales of numerous species of this family, and thus gain 
some idea of the range of modification and the general scale- 
characters of the group. In the synopsis which follows it 
‘will be seen that the subfamilies can usually be distinguished 
by scale-characters ; but the Bythidine are not sharply 
differentiated from some of the Neobythitine. Neobythites 
and Dicrolene, with numerous species, possess scales of a 
very uniform type, little modified from what we may suppose 
to have been the primitive pattern of the group. The tribe 
Bassozetini, with the radii absent or greatly reduced, must 
be looked upon as a specialized, not a primitive, group. 
This idea, based at first on the study of the scales, appears 
to be confirmed by the other characters of the fishes. The 
Lucifugine scale must be looked upon as a rather degenerate 
member of a primitive series. Brotula appears to have a 
primitive type of scale, but distinct in minor details, and not 
to be very closely associated with the Neobythitine. 
We are naturally led to ask, What characters have 
Brotulid scales in common with those of the Ophidiidz ? 
The general type of scale is quite the same, with coarse 
circuli, and numerous radii all around. There is no network 
Ann. & Mag. N. Hist. Ser. 8. Vol. xviii. 22 
