320 
BULLETIN OF THE BUREAU OF FISHERIES 
factor — that is, the supply of available food — -for this chaetognath is both extremely 
voracious and an active swimmer and hence would tend to gather at the levels, 
and probably to some extent to congregate in the regions where the copepods on 
which it chiefly preys are most abundant. Furthermore, it would naturally grow 
fastest and breed most actively where food was most plentiful, tending to produce 
and maintain an abundant local stock. 
It seems more probable that it is the dependence of S. elegans on the calanoid 
copepod plankton which, as remarked above (p. 30), is most plentiful in the mid- 
levels, which accounts for the comparatively sparse sagitta population of the deepest 
levels in the Gxdf of Maine and not the comparatively high salinity at these depths, 
for it thrives in still higher salinities in the North Sea region (Apstein, 1910). 
Temperature not only governs the distribution of S. elegans but also the size 
to which it grows, a fact that has long been recognized. Indeed, three varieties or 
subspecies of this species, one of them a large northern (“arctica”) , another a smaller 
boreal-temperate (“elegans”) , have been recognized by von Ritter-Zahony (1911); 
but Huntsman (1919) points out that these are not distinct, being connected by inter- 
mediates. In fact, the Gidf of Maine collections suggest that the difference in size 
between them probably is not hereditary at all, but the result of a direct physiological 
influence of the environment on the individual, for the adults average decidedly 
larger (up to 35 millimeters long) in March and April, when the temperature is near 
its lowest for the year, than in summer. This is not the maximum size for the Gulf 
of Maine, however, Huntsman (1919, p. 446) having recorded specimens of this 
length with ovaries still immature, and he describes S. elegans up to 52 millimeters 
long from the still colder waters of parts of the Gulf of St. Lawrence. He has also 
pointed out that it matures sexually at a smaller size in high temperatures than in 
low, as is the case with sundry other boreal planktonic animals — for example 
Aglantha digit ale. 11 
Sagitta serratodentata Krolm 
The fact that S. serratodentata is an annual immigrant to the Gulf of Maine and 
not endemic there has been brought out in an earlier chapter (p. 58), and its tropical 
origin and lines of dispersal have been discussed. It is safe to say there are no S. 
serratodentata in the inner parts of the gulf in late winter or early spring, the visitors 
of the previous summer all having perished, because our February and April 
cruises of 1920 did not yield it anywhere within the continental edge except for a 
single specimen in the southeastern part of the basin on March 11 (station 20064). 
It is probably to be found in the warmer water along the slope abreast of the gulf, 
however, throughout the year, for odd specimens were detected at our outer stations 
off the southwest face of Georges Bank on February 22 /station 20044), and off Cape 
Sable on March 19 (station 20077). 
In the year 1915 S. serratodentata had penetrated the eastern side of the gulf as 
far as the neighborhood of Lurcher Shoal and the northeastern part of the basin by 
May 10 (stations 10272 and 10273; Bigelow, 1917, p. 296), and by the last of that 
month and first days of June the Canadian fisheries expedition found it at two 
71 For a discussion oi other differences between the races of S. elegans living in high temperatures and in low see Huntsman 
( 1919 ). 
