192 
ME. ^Y. EAIEBAIEX AXD :ME. T. TATE OX THE DEXSITT 
series of readings was taken at intervals of three to five minutes, to obtain a reliable 
average, and the temperature again reduced. 
Thermometers. — It rvas found necessary' to maintain the temperature of the oil-bath 
considerably higher than that in the steam-bath, in order to make up for the radiation 
from the exposed portion of the tube and for the great retardation to the passage of the 
heat offered by the thick outer glass tube. The rule followed was to sustain such a 
temperature as would cause a gentle ebullition in the glass tube, trusting to the rapid 
convection of the water to secure a uniform temperature. Xo appreciable error could 
arise from this source from the smallness of the volume of steam in the globe-stem, and 
it was found in practice that a deposition of moisture almost immediately followed any 
cessation of the ebullition in the outer tube. 
For another reason the necessity of a higher temperatiu'e in the oil-bath became a 
source of serious difficulty. The difference between the temperature of the water and 
oil required to be constantly increased as the temperature and pressure became higher, 
because the radiation increased in a higher ratio than could be prevented by the use of 
ffannel coats, and a point was at last reached at which the imequal expansion of the 
thick glass tube on its cooler and warmer side, combined with the great pressime in its 
interior, caused a rupture attended by a dangerous explosion of great wolence. Twice 
with glass tubes, which should have borne a pressure of 500 lbs., and each time imder a 
pressure of only about 70 lbs. per square inch, an explosion occurred. In each case the 
fracture began in the bottom or closed end of the tube (unfortunately thickened by the 
glass-blower), as was evidenced by the boiler being lifted vertically from its seating, 
although the entire disruption of the tube immediately followed in the com-se of the 
explosion. 
Teiivperature reduced for ])ressure. — This column is calculated from the observed 
temperatures according to the ratio arrived at in the experiments ufith water pressime. 
The thermometers were made expressly for these experiments at the Kew Observatory, 
and the graduation was accurate within the limits of errors of observation. 
Capacity., in grains of water. — This column has been calculated from the readings on 
the cathetometer of the graduation marks, taken during each experiment. The capacity 
of the globe it is evident decreased as the inner column rose. 
Difference of lerel has been corrected to 32 °Fahr., at which temperatui-e one atmo- 
sphere of pressure is assumed to be equivalent to 29-9218 inches of mercui-y. 
Pressure of saturated steam, calculated from the experiments of M. Eegxault. — This 
is evidently the pressure of the steam in the steam-bath ; and if we deduct fi-om this 
the column of mercury in the globe-stem which this pressure balances, we arrive at 
the pressure of the superheated steam in the glass globe, given in the next column. 
Specific volume of the steam, obtained by dhiding the capacity of the globe at each 
temperature by the quantity of water introduced. 
In these experiments we were ably assisted by klr. W. C. Uxwix ; and we have much 
pleasure in stating that to his untiring zeal and superior skill in manipulation we are 
indebted for many important results. 
