354 BULLETIN OF THE BUREAU OF FISHERIES 



outlined above that it offered the opportunity for an important new fishery. 

 These early investigations likewise proved that it occupies a very definite environ- 

 ment, for it lives only along the upper part of the continental slope and on the 

 outer edge of the shelf where the inner edge of the Gulf Stream bathes the sea 

 floor as a band of warm water (47 to 50°) , and never ventures into the lower tem- 

 peratures on the shoaling bottom nearer land on the one hand, nor (so far as 

 known) downward into the icy Atlantic abyss on the other. Along this narrow 

 band it ordinarily finds the temperature not only warmer than the waters on 

 either hand, but varying by only a couple of degrees from season to season, and in 

 addition a bounteous supply of invertebrates to prey upon. But the balance 

 between the physiological nature of the fish and its surroundings is so delicate 

 that it lives in constant danger of disaster, and hardly had its range been mapped 

 when a submarine catastrophe overtook it. The first news of this disaster came 

 in March, 1882, when the master of a vessel reported dead and dying fish on the 

 surface, and throughout that month and the next vessel after vessel reported 

 multitudes of dead tilefish floating on the surface between the latitudes of Nan- 

 tucket and Delaware Bay. Digestion of all these reports ' outlined the area of 

 destruction as at the least 170 miles long by 25 broad — that is, an area of at least 

 4,250 square miles, and probably half as large again, thus covering the entire zone 

 inhabited by the tilefish north of Delaware Bay. At least a billion and a half dead 

 tilefish were sighted. 



It has generally been supposed (and we believe correctly) that the destruction 

 was caused by a sudden but only temporary flooding of the bottom along the warm 

 zone by abnormally cold water, consonant with which is the fact that other species 

 of fish suffered as well, and that dredgings carried on the following autumn proved 

 that the peculiar invertebrate fauna characteristic of this warm zone in previous 

 summers had likewise been exterminated. Unfortimately, however, no tempera- 

 tures were taken on the tilefish ground at the season when the mortality occurred, 

 and by the end of the following August the bottom water had again warmed to 

 48 or 49°. 



So complete, indeed, was the destruction of the tilefish that fishing trials car- 

 ried on off southern New England by the Bureau of Fisheries later in 1882, 1883, 

 1884 (when a particularly careful search was made and when the bottom water 

 along the tilefish ground was as warm as it had been in 1880 and 1881), 1885, 1886, 

 and 1887 did not yield a single fish.' But though decimated almost to the vanish- 

 ing point, the species was not quite extinct, as most people had come to believe, 

 the Grampus proving this by catching 8 off Marthas Vineyard in 1892. From that 

 time on the tilefish gradually reestablished itself, though the building up of the 

 stock must have been a slow process at first, for five trips and 18 sets of the line 

 trawl jdelded only 53 in 1893. Tilefish were next heard of in 1897, when a fishing 

 schooner caught 30 of 6 to 15 pounds weight on a haddock line trawl south of 

 Marthas Vineyard, and they had once more become so numerous by 1898 that the 



' Collins (1884b) has described the event in detail, as have many subsequent authors. An account will also be found in 

 Economic Circular No. 19 of the U. S. Bureau of Fisheries. 



' I have elsewhere summarized (Bulletin, Museum of Comparative Zoology at Harvard College, Vol. LIX, 1915, p. 237J the 

 temperatures taken in this region during the early years of the bureau. 



