INVESTIGATIONS OF THE ALBATROSS. 263 



which increase its length to a mile and a half. The rocks are all high 

 and bold, except one or two, which are but a few feet above water. 

 The island is covered with guano, the deposit of myriads of sea birds 

 which cover its heights, where they tind congenial quarters and safe 

 breeding- grounds. It is entirely devoid of vegetation, except a few 

 low leafless bushes. The only sign of vegetable life is a small patch of 

 grass a few feet in extent on one of the outlying rocks. 



A cave was discovered on the northwestern side of the island, into 

 which one of the ship's boats proceeded between 100 and 200 yards. 

 The arched roof was 15 feet or more in height at the entrance, increasing 

 as the cave was j)enetrated, the surface of the walls being quite smooth. 

 The water swarmed with fish. 



From Malpelo the line was continued about N. 36° E. true, in the 

 direction of the Gulf of Panama, where three normals were run from 

 deep water to the 100-fathom line, extending over two degrees of longi- 

 tude. The surface net was in constant use, and the Tanner net was 

 operated on several occasions. Serial temperatures were taken fre- 

 quently, the sounding wire being successfully used as a temperature 

 line. Several instruments were lost the first day by the parting of tbe 

 wire, but no losses occurred afterward, and the trouble arising from 

 the mercury shaking down was entirely eliminated. 



Reaching Panama for the second time on the morning of the 12th, 

 the ship was painted outside while repairs were being made in the 

 engineer's department. Tbe bunkers were filled with coal between 

 the IGth and 19tb, and we sailed at 1:10 p. m. on the 20th of March, 

 steaming across the gulf. 



Panama to the Galapagos Islands. — Reaching the vicinity of Galera 

 Point on tiie moridng of the 23d, a series of S(mndings was made, feel- 

 ing our way to deep water, as it vv^as desirable to cast the trawl and 

 make other investigations on the slope from 1,000 to 1,500 fathoms in 

 depth. Strong northeasterly currents were found in the vicinity, show- 

 ing a maximum of 51 miles in 21 hours, and continuing with varying 

 force nearly to the Galapagos. We were crossing the great Humboldt 

 Current, which sweeps along the coast of South America into the 

 Gulf of Panama. During nine months of the year, while the trades are 

 blowing steadily, the stream is divided, a portion entering the Gulf and 

 a broader belt taking a northwesterly direction. This division is not 

 usually apparent from January to May, while the trades are interrupted, 

 and the western part then takes a northerly direction, with its velocity 

 greatly diminished. 



The following series of observations extended from the South Ameri- 

 can coast to the Galapagos, crossing one of the great currents of the 

 Pacific Ocean at the point where its cooling waters pour into the Gulf 

 of Panama and where the faunas of the Southern and Central Pacific 

 meet. Great care was observed in all the work on this line and the 

 observations were varied and complete. The depths and serial tern- 



