CLUPE1TLE. 211 



the same locality in the middle of May their stomachs were crammed with sand 

 launces, Ammodytes, some of which were up to 2£ inches in length, and as many 

 as nineteen were inside one herring, while the sand-launces in their tnrn were 

 full of the remains of Crustacea. A month later from the same place, and 

 captured about eight mile3 off shore, the food had again changed, and consisted 

 almost entirely of the young of a very rare gobioid fish, Crystallogobius Nilssonii, 

 the largest of which was 1| inches long, there were likewise a few young herrings 

 and sand-launces. 



Goodsir (Edinburgh New Philosophical Journal, July, 1843) gives the result 

 of his examination of the Maidre, or food of the herrings in the Firth of Forth, 

 with which the stomachs of the fish were found to be filled. Cirripeds, Crusta- 

 ceans, and Acalepha were detected, of which crustaceans were in the largest 

 number, and consisted of masses of Amphipoda and Entomostraca. Among the 

 Acalepha the different species of Berooe were seen in the greatest numbers. 



Off Lofodens Sars found the sea swarming with microscopic animals, especially 

 small crustaceans termed " herring food," Calanidaz, chiefly Galanus finmarchius 

 and Temora longicomis. 



Carpenter writing upon what they consume in North America, remarks that 

 their food sometimes consists of smaller fishes, but generally of more minute 

 animals, especially Entomostraca and Radiolaria, of which small reddish-brown 

 aggregations are often found floating in the waters which they frequent. 



H. Widegren observed in Scandinavia that the food of the young as well as 

 of the grown herring consists chiefly of small crustaceous animals, invisible to 

 the naked eye, which are found in enormous quantities in the sea, both in shallow 

 and deep water, and may be detected by straining water through a straining cloth. 

 Their quantity varies at different seasons, and during a change of temperature at 

 different depths. Axel Broeck (Tids. Fisk. Kjobsch. i, p. 154, or Wiegm. Arch. 

 1862, p. 72) describes the various kinds of food on which he considered this fish 

 subsisted. Breeding herrings do not feed to the extent those not in this condition 

 do, but they do not invariably cease feeding, as some authors believe. 



In an aquarium it was found (Zool. 1876, p. 4856) that the young or 

 whitebait's diurnal and nocturnal movements were very dissimilar : in the former 

 period they were quiet and uniform in going round their tank in one shoal, but 

 at night the shoal was broken up, each fish taking an independent course for 

 itself, darting about from side to side, striking against the rock work with effects 

 fatal to themselves, and which was only stopped by placing a dim light to 

 illuminate the outline of the rockwork. Then they are most ravenous. 



Herrings live in an aquarium provided while being captured their scales are 

 not rubbed off, and Mr. Jackson suggests that the best plan is to dip them out of 

 the sea in a bucket or can, and subsequently allow them to swim into the tank 

 they are to occupy. He invariably observed that they swam round and round 

 their place of confinement in a circle from left to right. Occasionally one might 

 turn for a moment, but it speedily resumed its place, and he never perceived the 

 shoal to swim in a contrary direction. 



Migrations. — Many theories have been broached to account for the migrations 

 of the herrings, it having been formerly supposed that they yearly changed their 

 residence from the poles to southern waters. Even now some consider that it is 

 not, strictly speaking, migatory, that is, that it does not travel comparatively far 

 from the locality in which it was hatched, reared, and came to maturity, but 

 simply changes from shallow to deeper water, in accordance with temperature and 

 the supply of food. One proof advanced being that certain definite varieties are 

 present in certain waters. Where they conceal themselves is certainly remarkable, 

 thus along the Devonshire and south-west coast of England, Mr. Dunn observes 

 that should a gale spring up numbers are taken in nets purposely anchored parallel 

 to the shore, and that they are meshed on the land side, although this locality 

 had been unsuccessfully swept by seines and nets a very short time previously. 



That herrings, in common with other species of the family, occasionally 

 disappear from one locality, sometimes re-appearing in another, is well known, to 

 account for which many surmises have been made. From 1690 to 1709, a very 



14 * 



