272 
PACIFIC SCIENCE, VoL III, July, 1949 
macropterus, the identification of the species 
subsequently determined to be Parathunnus 
sibi proved to be most puzzling until an exami- 
nation of internal characters was made. It was 
possible, by inspection, to divide the catch of 
large tunas auctioned in the Honolulu fish mar- 
kets into two kinds, both of which, insofar as 
external characters went, agreed substantially 
with the available descriptions of Neotbunnus 
macropterus. It was suspected that one of the 
kinds was Parathunnus because of its large eye, 
large head, coarser scalation, and thick, heavy 
body. However, the available descriptions of 
Parathunnus differed in a number of respects 
from these fish. Although Kishinouye men- 
tioned that the pectoral fin was relatively shorter 
in larger fish, Kishinouye as well as Godsil and 
Byers described Parathunnus mebachi Kishi- 
nouye, which is here regarded as a synonym of 
Parathunnus sibi (Temmnick & Schlegel), as 
having a long pectoral fin reaching beyond the 
anal insertion and to or beyond the anal fin 
base. The Hawaiian fish examined by me have 
a pectoral fin which is shorter than that of 
Neotbunnus of comparable size and which 
usually does not reach as far as the insertion 
of the anal fin. Kishinouye described the anal 
finlets of P. sibi as grayish with a yellow margin; 
Fowler, who was familiar with Hawaiian ma- 
terial, reported that the finlets were without 
yellow markings. However, the finlets examined 
by me have been yellow or orange-yellow with 
black borders. A 100-pound specimen, the 
smallest available after several weeks of check- 
ing the markets, was purchased for dissection. 
A study of the internal characters of taxonomic 
importance- — such as the arrangement of the 
cutaneous circulatory system and the marginal 
striations on the liver — indicated that the fish 
was a Parathunnus , probably P. sibi (Temmnick 
& Schlegel). 
One of the most obvious differences between 
the Hawaiian specimens of P. sibi and the de- 
scriptions of this species in literature is, as has 
been mentioned above, in the relative length of 
the pectoral fin. Kishinouye’s Figure 47 (1923: 
PL 27) shows the tip of the pectoral fin reach- 
ing to a vertical line beyond the anal fin base. 
The photograph of one of the two specimens 
described by Godsil and Byers (1944: Fig. 59) 
shows approximately the same relationship. The 
pectoral fin in the Hawaiian Parathunnus, at 
least in the' size range examined by the Division 
of Fish and Game staff, hardly reaches a vertical 
line through the insertion of the second dorsal 
fin, and does not reach the anal insertion at all. 
There is an apparent difference in the rela- 
tive length of the pectoral fins of Hawaiian 
specimens and of species described in the liter- 
ature. This is probably attributable to the great 
differences in size range of fish examined in 
Hawaii and elsewhere. 
However, if the assumption is made that the 
relationship between the pectoral fin length and 
total body length is linear when logarithms of 
the body length are used, then the difference 
between the Hawaiian material and the avail- 
able descriptions may be reconciled. Since it 
seemed to fit his data best, Schaefer (1948) 
assumed a relationship of this kind between 
pectoral fin length and body length for Neo - 
thunnus . Though no proof of the assumption 
will be offered here for Parathunnus, it seems 
logical to assume that a similar relationship may 
exist. Where two variables are related linearly 
when the logarithms of one of them are used, 
then with an increase in the variables, the 
variable transformed into logarithmic form will 
increase much more rapidly on an arithmetical 
basis than the other. Hence as the fish becomes 
longer, the pectoral fin becomes relatively 
shorter, and conversely the smaller fish would 
have, therefore, relatively much longer pectoral 
fins. The size range of the Hawaiian Para- 
thunnus here reported was 1,191 to 1,900 mm. 
The specimen figured by Kishinouye (1923), 
and referred to as immature, was approximately 
750 mm. long as estimated by the scale indicated 
on the plate. The two specimens examined by 
Godsil and Byers (1944) were 569 mm. and 
910 mm. in length, respectively. A line fitted 
to pectoral fin length and the logarithms of total 
