302 BULLETIN OF THE MUSEUM OF COMPARATIVE ZOOLOGY. 
Bank with St. Croix. I found the connection perfect, but the ridge has 700 
fathoms water on it near St. Croix. There is 1,000 fathoms three miles north, 
and 1,800 fathoms five miles south of the ridge. I ran aline from Dog Island to 
White House Shoal, and back to Sombrero. Here I found a channel about 
ten miles wide, with 1,100 fathoms. "The temperature was 38° at 1,100 ; out- 
side 374° at 1,600, and 36}° at 2,500. I shall run a number of lines from 
St. Thomas to Sombrero, to be sure that this channel connects with the deep 
water off St. Thomas. I ran a line of soundings from the south end of 
Dominica to Avis Island. The soundings were regular at 1,000 fathoms, to 
within ten miles of Avis Island.” 
The soundings made by Commander Bartlett, after I left the 
“Blake,” to determine the ridges uniting the various islands between 
Sombrero and Trinidad, show plainly that the cold water of the Carib- 
bean can only come in through the passage between Sombrero and 
the Virgin Islands, which is about 1,100 fathoms, with a bottom tem- 
perature of 38°, while the 500-fathom line, as I have said, forms a 
gigantic island of all the islands to the south of Sombrero, including 
Dominica, with a narrow passage of 1,000 fathoms between it and 
Martinique; the 500-fathom line again uniting into one large spit, as 
a part of South America, all the islands to the south of it. Thus the 
bulk of the water forced into the Caribbean Sea has a comparatively 
high temperature, — an average, probably, of the temperature of the 300- 
fathom line. The cold water of the Atlantic is, however, again forced 
into the western basin of the Caribbean through the Windward Passage, 
and all this through the Yucatan Channel, between Cape San Antonio 
and the Yucatan Bank. It is, therefore, incredible that with this huge 
mass of water pouring into the Gulf of Mexico, there should be anything 
like a cold current forcing its way up-hill into the Straits of Florida, as 
has been asserted on theoretical grounds. The channel at Gun Key 
can only discharge the surplus by having a great velocity. 
Mr. Garman, who as usual &ccompanied mo, remained in the West 
Indies, after wo left the * Blake " at Barbadoos, for tho purpose of making 
collections of Reptiles and Fishes, with a view of throwing additional 
light on the former connections of the islands, as I have hore attempted 
to trace it. One of the most interesting of the Reptiles we collected is a 
gigantic land tortoise, found at Porto Rico, differing only in size from 
the land turtle still found on Trinidad and adjoining parts of South 
America. It is closely allied to the gigantic turtles of the Gallopagos, 
and to the fossil land turtles, of which fragments have been described 
by the late Professor Wyman. These were collected by Mr. A. Julien 
at Sombrero, in the phosphate beds of the island. 
CAMBRIDGE, May 10, 1879. 
