38 THE BOTANY OF BERMUDA. 
sustained high temperature of the winter, which is such that the tem- 
perature of the soil six inches below the surface was never found lower 
than 52°.66. 
Frost is nearly, but not quite, unknown. There are but two well 
authenticated cases on record. On the 24th December, 1840, while 
frost was visible ‘in low situations, water in tubs was frozen to the thick- 
ness of half a crown.”* On the 21st February, 1878, a thermometer 
on grass registered 28°.2. There is a tradition of snow, about 1811 or 
1812;+ a few flakes, indeed, fell at St. George’s on the 4th March, 1874. 
‘On 20th February, 187 2, the ground was in some places white with hail, 
which did not disappear for some hours. These facts are sufficient to 
Show that the temperature of the air is subject to much greater ex- 
tremes than are experienced at sea level within the tropics, although 
the vegetation partakes so largely of a tropical character. 
Low temperatures occur most frequently in the month of March. 
During the years 1872-77 a thermometer on grass registered below 
40° F., as follows: 
In December 2 times, lowest 35°.2 in 1876. 
In January 4 times, lowest 34°.0 in 1874. 
in February 3 times, lowest 359.0 in 1877. 
In March 8 times, lowest 35°.0 in 1877. 
Notwithstanding many years of observation, the data for determining 
the mean temperature of the air are imperfect. Observations have only 
been made at 9 or 95 a. m., and at 3 or 34 p. m., and the diurnal low for 
the several months is unknown. While the record was kept by the royal 
engineers, however, a monthly term day of hourly observation was ob- 
served, and from the days so recorded some approximation to the horary 
corrections might probably be derived, but I prefer to give the actual 
observations at those hours, subject to future reduction. To these are 
added in the subjoined table the mean temperature of the soil at 6 
inches and 12 inches depth, which will have future value, the Interna- 
tional Meteorological Congress held at Rome in April, 1879, having re- 
solved to include for the future the temperature of the surface of the 
earth among the meteorological elements to be observed. 
*Mr. J. L. Hurdis, in Jones’ ‘‘ Naturalist in Bermuda.” 
tMr. John Harvey Darrell is the authority for this statement. 
