December 9, 1910] 



SCIENCE 



843 



proportion of bunches only a part is living, 

 the dead parts being of a brownish color. 



Captain Dagnall's statements were con- 

 firmed by Captain George Morrison of the 

 R. M. S. P. steamship Berhice, who had been 

 in charge of steamers between Jamaica and 

 Southampton as well as of a steamer plying 

 between Jamaica and the Canaries. But he 

 thinks that patches of weed one acre in extent 

 are very rare, and he was unwilling to assert 

 that he had ever seen one larger than haK an 

 acre. In his opinion the gulf weed is torn 

 off from the Bahamas by the waves and the 

 greater part of it is swung around those is- 

 lands. The writer's own observations agree 

 with this, for in passing through the Bahama 

 archipelago along the seventy-fourth meridian, 

 he found the seaweed much more abundant 

 than along either of the lines followed across 

 the sargasso. The weed is evidently the same, 

 being in circular bunches up to 18 inches 

 diameter arranged in strips according with 

 the direction of the wind, though occasionally 

 in bands or even in patches 8 by 10 feet. The 

 patches are near the large islands. 



Seaweed occurs abundantly off the coast of 

 Venezuela. It comes from the borders of the 

 Orinoco delta and it was seen on the return 

 voyage at about west longitude 62°. Thence 

 it was very abundant to near the sixty-sixth 

 meridian, where it disappeared abruptly and 

 no more was seen except a small area near 

 the seventy-second. The abrupt disappear- 

 ance is difficult to explain; it is not due to 

 decay, for the last exhibition was of appar- 

 ently fresh weed; the distance from the source 

 is too small to justify the supposition of de- 

 cay, for the gulf weed is still living after hav- 

 ing traveled from the Bahamas round to the 

 east side of the sargasso. This is not the gulf 

 weed, to which it bears no resemblance. 



At best, the quantity of weed seen at any 

 locality is wholly insignificant. Midway in 

 the sargasso sea, the bunches seen in a width 

 of a mile would form, if brought into contact, 

 a strip not more than 65 feet wide. This, 

 where the weed is most abundant. But the 

 bunches are very loose, the plant material, as 

 was estimated, occupying less than one fifth 



of the space, so that if the bunches were 

 brought together so that the plant parts would 

 be in contact, each square mile would yield a 

 strip not more than 13 feet wide and 3 or 4 

 inches thick, or barely 2,500 cubic yards of 

 uncompressed seaweed to the square mile. In 

 most of the area traversed, the quantity would 

 be but a small fraction of 2,500 cubic yards 

 to the square mile — and the conditions are 

 the same along the lines described by Cap- 

 tains Dagnall and Morrison. The accumula- 

 tion of decayed vegetable material from sea- 

 weeds must be comparatively unimportant 

 under the sargasso sea; and what there is 

 would be merely foreign matter in mineral 

 deposits. 



The trade winds are comparatively gentle 

 in early July, and the latter part of August, 

 when the writer made the voyages; but they 

 are sufficient to raise waves of five to eight 

 feet and the sea is covered with " white 

 caps." Later in the season the winds become 

 much stronger. Reasoning a priori, one can 

 hardly conceive it possible that, with the 

 water in constant motion, the floating debris 

 could accumulate eii masse over any consid- 

 erable area; even should such accumulation 

 take place during a period of comparative 

 quiet so as to protect the water from wind, it 

 would soon be broken up by wave attacks along 

 the borders; for the patches of weeds are not 

 matted like peat — they are merely agglom- 

 erations of loose bunches drifted together. 



John J. Stevenson 



IS THERE DETERMINATE VARUTION ? 



Under this title I published in Science 

 four years ago' a paper discussing the changes 

 from year to year in the color pattern of the 

 beetle Diahrotica soror as these changes had 

 been observed by me during the decade 

 1895-1005. The observations depended on the 

 collection each year (1896-1900 omitted) at 

 approximately the same time and place (in 

 the later years two separate places each year) 

 of series of 1,000 individuals, and the determ- 

 ination and tabulation of the color-pattern 



'Vol. XXIV., pp. 621-028, Xovember, 1906. 



