10 
for older fish reaches its height during the latter part of May in 
this latitude. After spawning, the fish do not seem active for a 
time and as previously indicated, we find that migration far inland 
ceases almost entirely in New Jersey during about two weeks in 
August. 
When we consider that there are at least three kinds of fish 
of the same species’ coming in with the ‘tides toward the fresher 
waters, if is easy to explain on theoretical grounds, at least, such 
a condition as has been mentioned. 
In the very early spring we have the larger males and females 
which will spawn in a week or two; a large number of medium- 
sized fish which come in to feed but which will not spawn for a 
month at least; and lastly, the yearlings, which will not spawn until 
late in August or, if of the late brood of the previous year, probably 
not until the next summer. Knowing that after each fish is com- 
pletely spent of its genital products there must be a period of com- 
parative sluggishness, we find that the absence of the larger fishes 
from the marshes in August is readily explained; this also explains 
the relatively few medium-sized fish on the marshes in August. The 
fact that even the smaller fish do not appear in any great numbers 
in the middle of August, leads one to believe that another factor 
enters into the matter. 
This factor, the writer believes, is that of warmth. The fish 
which reacted positively to the fresher water as the temperature in- 
creased inland in the spring, may have reacted to the lower salinities 
on the marshes more readily when in a state of starvation. In the 
middle of August there are three conditions which might influence 
the return: first, the higher temperature which might reverse the 
reaction to fresher water; second, the condition of sluggishness fol- 
lowing spawning and inhibiting any battle with tides; and last, the 
fact that the animals are well fed and hence in a conditon to return 
to their abode for the winter. 
Early in September large numbers of Fundulus heteroclitus ret 
turn to the marshes with the tides, and they continue to run in and 
out, staying for shorter and shorter periods as the cold increases in 
the fresher streams, until finally they cease running until the follow- 
ing spring. 
The mating habits of Fundulus heteroclitus have been very well 
described by H. H. Newman (6) in the Biological Bulletin for 1907. 
The presence of males and females in aquaria together is known 
to be sufficient to cause the females to spawn. If a male is not pres- 
ent, when a gravid female is too much distended with eggs she will 
agstme a peculiar S shape, and vibrating her tail, will extrude some 
of her eggs. If a male is present, he will usually seek out and corner 
a female about ready to spawn and lock his dorsal and anal fins 
against hers. Then follows the simultaneous extrusion of eggs and 
sperm. 
