~ 
26 PHILOSOPHICAL SOCIETY OF WASHINGTON. 
the co-efficient constant C, as given by Prof. Ferrel’s theory, 
the two thermometers placed anywhere within doors or without 
would still give data for determining temperatures of the loca- 
tion; it should be borne in mind that the temperature thus ob- 
tained belongs specifically to the air in contact with the themome- 
ters and is not an average value for any extensive portion of the 
atmosphere. As it is an advantage to conduct observations under 
uniform conditions, it is recommended that a pair of bright and 
black bulb thermometers be attached to the whirling table, whereby 
the effect of a current of air may be on the one hand determined 
and on the other hand kept as ‘uniform as possible. 
Mr. Harxness said that the object practically sought by meteo- 
rologists was to learn the average temperature of a considerable 
body of air, but their efforts were thwarted by the irregularity and 
inconstancy of the distribution of temperature. So long as the air 
in contact with the thermometer is not precisely representative of 
the air of the vicinage it was useless to refine methods of observa- 
tion, unless by that refinement errors of a constant nature were 
eliminated. For the determination of mean monthly or annual 
temperatures he considered the reading of the nearest half degree 
as sufficient, and regarded the reading of the tenths of a degree as 
a useless refinement. . 
The advantage of reading to tenths was further discussed by 
Messrs. ABBE, DooxirrLeE, and Kummetyt. Mr. KumMMeE.y 
pointed out that where a difference of temperature is observed as 
an indication of the moisture of the air, the tenths are worthy of 
record. 
The following communication by Prof. Coartes E. Munrog, 
of Annapolis, Md., was then read by the Secretary : 
DETERMINATION OF THE SPECIFIC GRAVITY OF SOLIDS BY THE 
COMMON HYDROMETER. 
Having occasion some time since to devise methods for the ex- 
amination of coal on board ship, I was obliged, as my first con- 
sideration, to work with such materials and apparatus as are usually 
found in ships’ stores, and then to arrange the methods so that they 
could be used under the restricted conditions which prevail. The 
unsteadiness of the ship makes balance methods for the determina- 
tion of specific gravities difficult, even when a suitable balance is at 
