406 Messrs. W. Kellner and AY. H. Deering. [Mar. 7, 



rather some oxygen and carbonic oxide would co-exist), while the 

 quantity of heat is measured in the cooled-down gases when all 

 oxygen and carbonic oxide have combined ; here, also, it would not 

 be practicable with existing information to make an accurate allow- 

 ance for difference of heat-quantity tinder the conditions of the 

 experiment. Further, in calculating the pressure of the gases from 

 their calculated temperature, it is assumed that their coefficients of 

 expansion which were determined for to 100 C. are valid at 

 3000 C. (At 0, air is only about 190 above, and water is 100 

 below, its boiling temperature.) 



If temperature were independently known, the drawbacks just 

 mentioned in the use of the coefficient of expansion, and of the 

 volume of gases, still apply to the calculation of pressure ; and, 

 conversely, to the frequent case of calculating temperature from 

 ascertained pressure. 



The steel explosion- vessels used in our experiments were of about 

 120 c.c. capacity ; they were closed by screw stoppers of about 

 4 ins. total length (the screw portion about 1-J ins. long), the end of 

 which was screwed down to gas-tight contact with the circular seat- 

 ing forming the mouth of the explosion-chamber. The stoppers had 

 a square head at starting 2 ins. long ; the whole stopper was per- 

 forated axially with a cylindrical hole J in. in diameter. The mouth 

 of the hole in the stopper was closed by a steel ball of ^ in. in diameter 

 (which formed the valve), a ball such as is used for ball-bearings; on 

 the ball a 1-in. thick iron plate, about 15 ins. by 15 ins., was placed 

 (the centre of the plate being perforated by a cylindrical hole J in. in 

 diameter, resting on the top of the ball), and on the plate the weights. 

 The latter consisted of lead cylinders, weighing about 5 cwt. and 

 8 cwt. respectively (diameters 14 ins. and 12 ins.) ; of the former, at 

 the most three (placed on top of each other), and of the latter, two,, 

 were used. The weight was made up with half-hundredweights and 

 smaller weights placed on the top of the upper lead-cylinder ; all the 

 weights, lead-cylinders, iron plate, and iron rod, were weighed at the 

 beginning and end of the experiments by standard weights. 



The explosion- vessels had externally a collar and a hexagonal por- 

 tion fitting into a hexagonal hole in an iron plate, which was screwed 

 to a balk of timber, held in place by weights ; this held the explosion- 

 vessel while the screw-stopper was hammered down, and formed the 

 base of the apparatus. The lead cylinders had an axial cylindrical 

 hole to their centre ; in the hole of the top cylinder an iron rod 

 (about 1 in. diameter, 4 ft. long) was placed, the upper end of the 

 rod passing loosely through a hole in a cross-beam kept in position 

 by iron uprights on either side of the base of the apparatus, The 

 weights were carefully balanced, so that pressure of iron rod against 

 cross-beam was reduced to an inconsiderable amount. The Lead 



