Munroe: Systematics of western Atlantic Symphurus 
69 
143/175 (82%) of the individuals were captured. Of 
interest is the capture of a 27.3 mm specimen at 329 
m, and a 38.5 mm specimen at 238 m, indicating that 
small juveniles also occur at depths inhabited by 
adults. Only 5/175 (3%) S. piger have been collected 
at depths shallower than 110 m, the shallowest cap- 
ture (92 m) being that of three specimens (UMML 
35255) taken off the Netherlands Antilles. Only 21/ 
175 (12%) fish were taken deeper than 300 m, with 
the deepest record (UMML 7124) being that of four 
specimens taken at 549 m off Puerto Rico. 
Remarks In his revision of western Atlantic tongue- 
fishes, Ginsburg (1951) stated that he was unsure 
which, if any, of specimens on which the original de- 
scription was based, was the holotype of A. pigra. 
He selected ( 1951:197) the specimen from Blake Sta- 
tion XXIII (now MCZ 27965) as the lectotype. This 
action was invalid, because in the original descrip- 
tion Goode and Bean (1886:154) designated the speci- 
men from Station XXIII as “the type” of A. pigra and 
listed specimens from Albatross Stations 2318 and 
2405 as collateral types (^paratypes in current ter- 
minology), not cotypes as stated in Ginsburg ( 1951). 
This distinction is important because of all presently 
available specimens believed to form the basis for 
the description of A. pigra, the only specimen that is 
actually this species is the holotype collected at Blake 
Station XXIII (see below). 
Among seven type specimens included in the origi- 
nal description of A. pigra (Goode and Bean, 1886), 
all still extant, with the exception of the holotype, 
are S. parvus Ginsburg, 1951. In addition, among 
other, nontype material listed in the original descrip- 
tion, Goode and Bean included specimens of a third 
species, S. minor Ginsburg, 1951. The status or ex- 
istence of this material, other than the holotype, is 
presently somewhat confusing. Ginsburg ( 1951) and 
I have tried unsuccessfully to locate all specimens 
included in the original description of A. pigra. Diffi- 
culty with tracing this material results from the fact 
that specimens in the original description were listed 
only by Albatross or Blake station numbers. Goode 
and Bean, furthermore, did not always provide the 
number of specimens examined from each station. 
Although some tonguefishes from these stations 
were located, it is uncertain whether these are the 
specimens used by Goode and Bean. For example, 
Ginsburg (1951) was unsure that specimens he ex- 
amined from Albatross Station 2318 were the ones 
used by Goode and Bean in the original treatment of 
A. pigra. Ginsburg was also unable to locate two 
specimens from Albatross Station 2405 and an un- 
known number of specimens from Station 2425 listed 
in the original description of this species. It is note- 
worthy that Ginsburg discussed not four specimens 
from Albatross Station 2318, as in Goode and Bean’s 
original description, but rather six specimens. Ap- 
parently, Ginsburg was unaware of the discrepancy. 
During the course of the present study, attempts 
were made to locate the six paratypes of A. pigra 
and to discover the reason for the discrepancy in 
specimen number. Field data now included with these 
six USNM specimens reveal that all jar labels and 
museum registers list data only for Albatross Sta- 
tion 2318. If station data for all six specimens are 
correct, Goode and Bean either erred in listing only 
four specimens from this station (not likely), or two 
additional specimens collected at this station were 
not included in the original description (possible). If, 
however, the four specimens from Albatross Station 
2318, as reported in Goode and Bean’s account, are 
the correct number, and these authors used all avail- 
able material, then station information for two of the 
six specimens now listed from Alba tross Station 2318 
have been lost or transposed prior to examination 
by Ginsburg. Efforts to retrace the history of these 
specimens prior to their inclusion in USNM holdings, 
including a check of museum registers and catalogue 
of the Bureau of Fisheries Collection at the USNM, 
uncovered no additional information. A search 
through accession files at the USNM regarding trans- 
fer of these specimens from the Bureau of Fisheries 
also was unsuccessful because no accession num- 
bers) were listed for the specimens. Therefore, I can- 
not unequivocally demonstrate that the discussion 
below regarding the history of paratype material for 
A. pigra is fully correct. 
The six specimens listed in Ginsburg (1951) sup- 
posedly collected from Alba tross Station 2318 are now 
assigned the following museum numbers: four are 
USNM 74330 (three were designated as S. parvus 
paratypes by Ginsburg; the fourth, not listed in 
Ginsburg, was on loan at the time to R Chabanaud 
( MNHN ) is also S. parvus but was not designated as 
a paratype); the two other specimens from Albatross 
Station 2318 also are part of the type series of S. 
parvus. One specimen is now USNM 84491, the ho- 
lotype of S. parvus and the other (USNM 152733) is 
a paratype of S. parvus. Both specimens originally 
were contained in the same jar (indicated both in mu- 
seum catalogue records and on jar labels). These 
(USNM 84491 and 152733) may be the two specimens 
from Albatross Station 2405 listed in Goode and Bean’s 
account of A. pigra. If this assumption is correct, then 
these specimens together with the four in USNM 74330 
would account for the six paratypes of A. pigra. 
The fate and status of the remaining material listed 
in the original description of A. pigra is as follows. 
Specimens from Albatross Station 2374 (nontype sta- 
