528 U. S. BUREAU OF FISHERIES 
Island to Gloucester, most of them to the southwestward of the place 
of release. The tagging operations have been too limited to warrant 
any definite conclusions. It is hoped to continue them in 1927, when 
a more suitable tag may be devised. 
SMELTS OF NEW ENGLAND 
A general account of the natural history, fish-cultural propagation, 
and conservation of,the Atlantic smelt, with a history of the smelt 
fisheries, by Dr. William C. Kendall, to which reference was made in 
a previous report of this division, has been published. That account 
is being supplemented by another paper on the smelts of the genus 
Osmerus, which embraces life histories, age, rate of growth, racial 
peculiarities, etc. The original plan to include Pacific species of 
the genus has been simplified by a recent reviewer of Pacific smelts, 
who has relegated all but one species (formeriy regarded as Osmerus) 
to other genera. Work on the smelts is discussed at greater length 
on page 614. 
MULLET 
The gray-mullet (Mugil cephalus) investigation in the south 
Atlantic States was continued during 1926 at Beaufort, N. C., by 
Elmer Higgins and Robert O. Smith. The tagging experiment was 
extended, 1,000 fish being tagged and lberated at Beaufort during 
July, August, and September. Because of the difficulty of securing 
uninjured fish in quantity from commercial hauls, a beach seine was 
operated by the investigators and crew. ‘This made possible the 
tagging of mullet and at the same time the tracing of the growth of 
mullet spawned in the preceding fall. Due to various factors, 
among which might be mentioned more effective advertising, smaller 
commercial catch, and the different sizes of fish tagged, 34 tags were 
returned during the year, an average of 3.39 per cent, or one returned 
for every 29.5 fish tagged. In some cases, 2 to 5 individuals from the 
same school were returned. All tags were received from North 
Carolina, except two from the upper portion of the South Carolina 
coast, thus confirming the results of previous experiments. ‘There 
were no returns from fish tagged in 1925. 
The assumption that so-called “ Cape mullet ” constitute a distinct 
division of the local stock seems to be corroborated by analyses of 
commercial catches during 1926. These fish appear at Beaufort 
about the middle of September, and are plentiful until the first of 
November. Length-frequency studies of Cape mullet that were 
taken in a number of commercial hauls indicate that these fish are 
intermediate in size between the local O and I classes. However, 
during the latter part of this season (mid October to mid November) 
it was virtually impossible to separate local O-class and “Cape” 
mullet on the basis of average length, inasmuch as some of the com- 
mercial Cape mullet averaged extremely small in size. 
Study of the growth of mullet, particularly the O-class, was con- 
tinued and significant results were obtained. Juvenile mullet of 22 
millimeters body length appear at Beaufort by the middle of Jan- 
uary. They become more abundant during February, March, and 
April, but there is little growth until the middle of April, when the 
water temperature reaches about 20° C. In May the rate of growth 

