560 
BULLETIN OF THE BUREAU OF FISHERIES 
paratively common, whereas, M. eulepis is very rare, as previously stated, it seems 
logical to refer all the young, at least tentatively, to holmesi. 
The locally represented species of Microgobius reach a length of only about 2 
inches and they probably are of only slight economic value even as forage fish, because 
of their comparative scarcity. The rather large number of young taken suggests, 
however, that the fish may be somewhat more common than is indicated by the 
scarcity of adults in the collections. 
Neither adults nor young were taken from December to February, inclusive. 
It seems probable, therefore, that these fish leave the local waters during the winter, 
or that they possibly seek shelter in the mud or sand like Fundulus and probably 
other minnows. 
SPAWNING 
The eggs of Microgobius have not been studied. M. holmesi with large roe 
were taken only during the first half of July. That the spawning period of this 
species is not limited to the month of July is evident, however, from the collection 
of larvae at hand, as shown subsequently. Smith (1907, p. 368) reported that a female 
M. eulepis distended with nearly ripe eggs was taken at Beaufort on May 18 (1905). 
No gravid fish of the last mentioned species were seen during the present investigation. 
A few young Microgobius, only about 3.0 to 4.0 mm long, were taken as early 
as March 11 (1929), and a few equally as small were taken as late as November 21 
(1927). The larvae were numerous, however, only during July, August, and Sep- 
tember. The collection of young Microgobius indicates, therefore, that some spawn- 
ing takes place as early as March, that it continues throughout the summer, probably 
extending into the month of November, and that the principal spawning season 
occurs during July, August, and September. 
Larval Microgobius were taken over the entire area in which tow-net collections 
were made, including Beaufort Harbor, the adjacent sounds and estuaries, and off 
Beaufort Inlet to Cape Lookout and as far as 12 to 13 miles offshore. It seems 
reasonable to expect Microgobius to produce eggs which become attached like those 
of Gobiosoma and Gobionellus, and like those of the various European species that 
have been studied. If that be true the eggs do not drift and the recently hatched 
young should be expected to occur somewhere near the place where the eggs were 
spawned. It seems probable, therefore, that spawning takes place over much of the 
area in which the larvae were taken. 
BESCRIPTIONS OF THE YOUNG 
Specimens 1.6 to J+.O mm long . — Specimens of Microgobius 4.0 mm and less in 
length generally are difficult to separate from Gobiosoma (figs. 50 and 51). A careful 
study has revealed no outstanding structural differences in these small larvae, and 
color marking in preserved specimens, except in rather rare instances, are not of much 
help until a length of about 4.0 mm is attained. In general, the vent in Microgobius 
is slightly more anterior in position than it is in Gobiosoma. This difference is evident 
only when specimens of even size are compared, and it is not readily shown in a table of 
measurements, because few specimens are straight enough for an accurate measure- 
ment and, furthermore, the position of the vent evidently changes with growth. The 
quotients derived from dividing the caudal portion of the body into the total length to 
the tip of the notochord, even in specimens varying less than a millimeter in length, do 
not show a constant difference. An average difference, however, is evident, for in 12 
specimens measured of each genus, the caudal portion of the body averages 2.15 
