REPRODUCTION AND DEVELOPMENT OF SCIAENIDAE 
103 
as already indicated, leaving the shallow waters in the autumn or early in the winter 
and returning early in the spring when low temperatures are not likely to recur. 
Although the smaller gray trout generally migrate away from the shallow waters 
during cold snaps, they apparently do not always go early enough, for on January 14, 
1927, quite a few dead and numb fish were seen along the shores of Pivers Island. 
A sample of 6 fish was gathered having a range in length of 122 to 182 millimeters. 
Several days of rather unusually cold weather preceded January 14. The water 
temperature at the laboratory pier, taken daily at 4:30 p.m., dropped to 41° F. on 
January 9, and it remained there until January 14 when it came up to 48° F. A 
temperature of about 41° F., therefore, probably is close to lethal. 
SPAWNING 
The spawning season of the gray trout is a moderately long one. Fish with 
well-developed roe (although not ripe) were observed as early as April 16 (1914), 
and as late as August 17 (1914). The young were taken first on May 25 (1932). 
However, they are rather large ones, ranging from 6.5 to 9 millimeters in length, 
and therefore may have been a few weeks old when caught. Larvae under 5 milli- 
meters in length were caught during June, July, and as late as August 27 (1930). 
According to these data the spawning season at Beaufort begins in May and extends 
through June, July, and August. 
The conclusion relative to the spawning period at Beaufort is in general agree- 
ment with statements by Welsh and Breder (1923, pp. 150-158), who say that the 
season is a protracted one, commencing in May and continuing until September. 
The great majority of the fish are reported, however, to spawn between the middle 
of May and the middle of June. These authors state, furthermore, that the spawning 
season is little affected by the latitude and that it occurs at about the same time from 
Cape Cod to the Carolinas. Higgins and Pearson (1927, p. 57) found virtually all 
mature females taken in pound nets in Pamlico Sound, N.C., in a spawning condition 
in the second week of June (1925) when they began their investigation, and by 
August 10 spawning had been completed. 
It was noticed at Beaufort that the large fish develop roe earlier than the smaller 
ones. Among thousands of fish examined only large ones had gonads in an advanced 
state of development in April, whereas in August those fish that still contained roe 
were small ones, generally 15 inches and somewhat less in length. 
In the vicinity of Beaufort spawning apparently takes place principally, if not 
wholly, at sea. This conclusion is arrived at from a study of the movements of adult 
fish and from a study of the distribution of the young. It has been stated elsewhere 
that the large fish return to the inside waters early in the spring from as yet a rela- 
tively unknown winter home. For a time these fish are quite abundant. However, 
within a month or two after their arrival, or about the latter part of May, the large 
fish become scarce and remain so until about July when they again increase in abun- 
dance. During two seasons (1914 and 1915) of almost ceaseless effort no females 
quite ripe were found among thousands of fish examined. Virtually all contained 
roe early in the spring. Later, when the large fish returned after a period of absence 
they had spawned. The explanation appears to be that the fish go out to sea to 
spawn and return to the estuaries and sounds when the process is completed. 
The foregoing explanation of the scarcity of large fish in the inside waters during 
a period, which undoubtedly is the height of the spawning season, is substantiated in 
57094—34 5 
