234 
BULLETIN OF THE BUREAU OF FISHERIES 
length. Except in occasional pools the brook is nowhere over 3 feet wide in ordi- 
nary height of water, until within two or three hundred yards of extreme high-tide 
mark. At this distance above high-tide mark there is an old stone bridge, through 
which the water flows over large rocks, making a cascade beyond which smelts could 
not pass. The brook then becomes considerably wider, with a few pools and inter- 
vening shallow ripples over gravelly and rocky bottom. Below high-tide mark the 
stream becomes a tidal creek, with sandy bottom in the channel or continuation of 
the brook, and with mud flats and marsh on either side, which are covered by salt 
water at high tide. This condition extends for about a quarter of a mile, when it 
broadens out into the head of Harraseeket River, a tidal arm of Casco Bay, at “Porters 
Landing,” Freeport, Me. At the lower end of this brook proper there is a short 
extent that salt water covers only during high runs of tide. At other times the 
water is unaffected by the tide. The immediate banks of this portion of the 
stream are without bushes or trees, but above, as far as the stone bridge, it is mostly 
thickly overhung by alders. 
Just below the lower end of the creek another much smaller and shorter tidal 
creek comes in on the right. Its channel is very shallow and muddy below the 
fresh-water brooklet that supplies the fresh water. The brook is a mere spring-fed, 
woodland rivulet, which in narrow portions is not over a foot wide but in places is 
relatively deep, especially in the marsh near high-tide mark. To distinguish it from 
the previously mentioned brook, the boys used to designate it as the “Little Brook.” 
Smelt ascended Little Brook also, but according to the present writer’s recollections 
not as early as in the large brook. Why smelts should divert themselves into this 
little creek and brook when their course was practically straight away up the larger 
stream is hard to explain. 
As soon as the ice was out of the creek and the water of the brook fairly clear 
the smelts would appear. The time of the first appearance varied more or less, 
according to the season, from the latter part of March to some time in the first 
week of April. 
There are other brooks in the region in which usually smelts did not run quite 
as early, being considerably larger streams and requiring a longer time to become 
clear of ice and turbidity. 
The first smelts to appear were comparatively small, dark-hued males, their bodies 
rough with tubercles, which, when a fish was taken in the hand, made it feel as 
though covered with sand. Later the smelts would run in gradually increasing num- 
bers, comprising both sexes, until in the latter part of the season it would hardly be 
exaggeration to say that the brook was full of them. It is literally true that on some 
mornings after a big run the pools would be black with smelts, as, looked at from 
above, the fish appear black or very dark. 
The runs are known to extend well into May, but gradually decrease in number 
near the last of the season. No intermittent runs, as mentioned of the smelts in 
Massachusetts, were noticed, although fish would appear more abundant at certain 
times than at others. Of course, there might be times of heavy rains and freshets 
that prevented runs of smelts, or at least they could not be seen if present. 
Smelts never were known to run in these brooks in the daytime, or, so far as any 
authentic information goes, in any other brook in this vicinity, and at no time did 
