> 
Scales of the Brotulid Fishes. 325 
all around, no break between the lateral and basal series ; 
apical field with a fine network between the circuli. WN. lon- 
gipes, Smith & Radcliffe, N. purus, Smith & Radcliffe, and 
N. fasciatus, Smith & Radcliffe, have exactly the same 
pattern as N. steaticus. 
Dicrolene scales (D. longimana, Smith & Radcliffe; D. tristis, 
Smith & Radcliffe, D. intronigra, Alcock) do not differ 
from those of Neobythites. 
D. intronigra is remarkable for the length of the scales, 
which are about 4°25 mm. long and 2 broad. In D. tristis 
they are about 3°75 mm. long. Many of the scales of 
Neobythites and Dicrolene are latiuucleate. 
A note should be added with reference to Bregmaceros. 
In Proc. Biol. Soc. Wash. xxvi. p. 77, I was certainly in 
error with reference to the ‘“squarish to suboval plates,” 
which appear, in good dry material, to be really depressions. 
They, therefore, have nothing to do with the veritable trans- 
versely ridged plates of Gadus, the Ophidids, and Brotulids, 
which really are distinct elements, and in some mounts can 
be seen as distinct entities, separate from the scales to which 
they were attached. Bregmaceros japonicus (Misaki, Japan ; 
Stanford University) shows exactly the same features as 
B. atripinnis, except that the very broad scales have no 
angulation of the circuli in the middle line above. 
Aside from the question of Bregmaceros, it now appears 
certain that the theory of Tims, that scales arose from the 
fusion of separate placoid-like elements, is incorrect. In the 
remarkable shark Hridacnis radcliffei, Smith, the scales are 
arranged in rows exactly as in Teleosts ; and although they 
are of course placoid, with a central spine, they are broadly 
expanded laterally, and the expanded portions have a minute 
reticular sculpture. “The sculpture on the EHridacnis scale 
is, however, represented by lines enclosing hexagonal areas, 
which are more or less narrowed at either end, and are 
placed so that the middles of those of one row approximately 
correspond to the ends of those of the rows adjacent. There 
is here a distinct resemblance to the sculpture of Ceciliid 
(Amphibian) scales. Granting that placoid scales are 
exceedingly different from those of Teleosts, as shown by 
Goodrich and others, they, nevertheless, surely represent 
essentially homologous structures. 
