and Growth-rate of the Lesser Sand-Eel. 285 
October.” No proof whatever is given for this statement ; 
but an age of five months is assigned quite arbitrarily to the 
March and April larva’, in the face of their very “ young” 
characters, and from this it is supposed that the spawning- 
period must be assumed to last till October. “The,” as he 
remarks, ‘“ purely conjectural estimate” of five months as the 
age of the larvee—although disagreeing with all the facts and 
quite a needless over-estimate—is taken as a proof’ of the 
second assumption :—“ ‘The duration of the spawning-period 
... may therefore be set down as from the end of June till 
October.” 
Following up the young forms of March and April, we 
find a complete transition in sizes, and the growing sand-eel 
may be traced from the bottom into the mid-water and to the 
surface. In Table II. the supposed quiescent period in the 
sand is represented by longitudinal shading, and the average 
duration of the bottom-residence is indicated by dotted lines ; 
the passage through the mid-water is indicated by dots and 
lines, and the surface-habit by transverse lines. The two 
limiting lines of the curve are plotted from the limiting sizes 
which occur in the collection, set forth in Table I. If the 
upper line were exactly parallel to the lower, then the growth- 
rate of all the fish would be exactly equal—the fact that they 
are so closely parallel speaks strongly in favour of the theory 
put forward in a former paper (plaice) that the difference in 
size shown in “hauls” of larvee is due directly or indirectly 
to the prolonged spawning-period. The earlier-hatched 
forms, with a slight superiority in competition, show a little 
faster growth, so that the lines diverge slightly. It will be 
observed by comparing ‘Tables I. and IL. that this March 
curve, up to some 55 millim., is not hypothetical except in 
a very slight degree, the points on the curve being plotted 
out in accordance with a great number of forms, only a few 
leading ones being indicated by crosses. ‘The second and 
third year curves must be conjectural, but at any rate offer an 
explanation agreeing with the facts. 
Coming to the July forms, we have the same occurrence in 
few numbers of the small post-larval sand-eels, which, taken 
in conjunction with their absence in May and June, is strong 
evidence in favour of the occurrence of two spawning-periods. 
In this case the first larvee occur in the third week of July, 
and a larval period of two weeks, added to an incubating 
period of ten days, takes us back to a spawning-date of 
the latter part of June, which has already been observed 
by M/‘Intosh and confirmed by Fulton (see above). The 
only difficulty one has to overcome is that there is not appa- 
