122 
BULLETIN OF THE BUREAU OF FISHERIES 
weaker. In fact, it may not be followed at all there, for this pteropod is occasionally 
met with in great shoals on the surface off Iceland in daytime, though usually not 
when the sun is high. 
The presence of Limacina retroversa in the Gulf of Maine throughout the year, 
together with its very general distribution there, proves that its local presence or 
absence is not governed by small variations in temperature or salinity. On the 
contrary, Limacina (both large and small) has been taken at one season or another 
in water varying in temperature from 2° to about 16.6° — that is, over practically 
the entire range proper to the gulf except for the very coldest and the very 
warmest. Probably its habit of coming up to the surface at night brings it into 
the latter also, on occasion. But the great majority of the Gulf of Maine records 
for this pteropod have certainly been from temperatures lower than 15° at all sea- 
sons, and since it has never been found regularly or abundantly in water warmer 
than this in any part of the ocean, 15° may be set arbitrarily as the upper tem- 
perature limit to its continued presence and prosperous existence. Thus, in our 
latitudes it is probably the high temperature of the oceanic water that is the offshore 
barrier to it, confining it to the continental edge and shelf off the coast of the 
United States. 
On the other hand, although Limacina occurs in temperatures as low as 2 to 3° 
in the gulf in winter, it does not tend to congregate in the very coldest water at 
that season, but rather the reverse, for it was either absent altogether or at least 
very rare during the spring of 1920 (one or two only at stations 20055 to 20061) 
wherever the major part of the column of water was colder than 2°, although it 
was present in the neighboring parts of the gulf at the time. We have found it 
equally lacking or very rare in early spring in the icy cold water over the whole 
breadth of the shelf abreast of southern Nova Scotia, and certainly it is very scarce, 
if it occurs at all, in the coldest water along that coast in summer. Furthermore, 
Doctor McMurrich’s notes show that there is a very close agreement between winter 
chilling and scarcity, vernal warming and regular presence of Limacina at St. Andrews, 
where it practically disappears when the temperature falls below about 3°, not to 
reappear regularly in the tows until the water warms to 8 or 9° the following spring. 
Although the evidence is not so clear, it seems that the presence or absence of 
Limacina may be correlated similarly with temperature in Massachusetts Bay, 
whence it appears to vanish when the water chills below, say, 2 to 3°, as happened 
in February and March of 1920; whereas in warmer winters, as that of 1912-1913, 
when the temperature of the water did not fall much below 3°, Limacina may 
occur sporadically and in small numbers right through from autumn until February 
(p. 120). These facts obviously suggest that it is the local cooling of the water that 
drives this pteropod from the coastal waters of the gulf, and from its northeastern 
corner generally, in late winter and early spring. 
Temperature may also determine the bathymetric occurrence of Limacina. 
For example, we found it comparatively abundant on the surface over the outer part 
of the shelf abreast of Cape Sable early in the summer of 1915 (station 10294, 
June 23), when the superficial water had warmed to 9° to 10°, but with temperatures 
as low as 2° to 3° only 40 meters down it was certainly scarce at deeper levels. In 
