Marshall—The Vegetation of Twin Island. Ill 
rose above 60° for a week, the maximum being 63°. During 
the-first half of July, then, it was about 11° below the water. 
During the latter half of July when the temperature of the 
water remained at 1 1°, the back strand was below 62°, that is, 
the difference between land and water was relatively less. For 
the first part of August the temperature ranged between 58° 
and 60°, and the difference between land and water was still 
less. When the water reached the maximum, in the second 
week of August, the land likewise rose in temperature, regis¬ 
tering between 67° and 72°, although not as rapidly as the 
water. Then when the temperature of the water fell and re¬ 
mained below 72° for the rest of August, the land did not 
fall below 60°. The land had a temperature a few degrees 
higher at the end of August than at the beginning of July. 
The lowest temperature recorded was 55°, the highest 72°. 
Observations taken from the middle of July to September 1 at 
another station a few feet away but about one meter nearer 
the now receding water, showed a temperature almost uni¬ 
formly one degree higher. 
The temperature of the soil of the back strand remained 
always below that of the strand, except for one day; however, 
the records were made for a greater depth, one foot. The 
temperature showed a slight rise during July; it reached a 
maximum of 56° at the time when the water reached its max¬ 
imum for the month. The temperature then remained close 
to 56° until the second week of August when it reached a 
maximum of 66°, when all other temperatures were highest, 
and did not fall below 61° until the last two days of August. 
For the last week it remained only about one degree below the 
soil of the strand. 
The temperature of the soil at the station in the center of the 
island among trees and shrubs taken at the same depth as the 
back strand temperatures, one foot, did not vary more than four 
degrees from the latter. For a few days the soil of the center 
reached this upper limit. In the first half of the summer, the 
temperature remained much nearer that of the back strand 
soil; occasionally it registered one to three degrees below it. 
When the general rise occurred in the second week of July, the 
