May 20, 1880] 



NATURE 



63 



changes in volume. The latter (Figs. I and 2) consisted 

 of a massive block of cast iron containing two cavities ; 

 one (c) for the reception of the extremity of the column 

 of mercury, the second (d) for the reception of the gradu- 

 ated tube (iM) inclosing the gas to be experimented upon. 

 A narrow passage connects the two with each other 

 and (f) with the reservoir of a powerful pump, while 

 conical screw-taps (P', p") manipulated from the outside 

 permit the openings into the cavity beneath the column of 

 mercury or into the reser\-oir of the pump to be closed at 

 will. The manometer (m) containing the gas to be com- 

 pressed is cf glass tubing, having an internal diameter of 

 I miUimetre and an external diameter of 10 millimetres, 

 and is graduated for a distance of 50 centimetres. It is 

 inserted hermetically into a massive bolt (b), which enters 

 into the second cavity (d) of the apparatus. The free 

 portion of the manometer is inclosed by a roomy glass 

 tube, through which flowing water maintains a constant 

 temperature, and that in turn by a copper cylinder, to 

 guard against accidents. Mention can only be made 



here of the ingenious devices for closing hermetically all 

 the joints. The second important feature, the tube (c) 

 for containing the column of mercury, is formed of pieces 

 of steel tubing — internal diameter 2 millimetres, external 

 5 millimetres — united by specially prepared joints, which, 

 while closing hermetically, are still easily attached or 

 detached. The place chosen for the most noteworthy 

 experiments was the , coal-mine of \'erpilleux, in the 

 neighbourhood of St. Etienne. This pit reaches a depth 

 of 327 metres, and a constant temperature prevails at the 

 bottom. 



The experiments made here were confined to nitrogen 

 gas. The chief features of an experiment are as follows : — 

 A vessel containing svarm, dry mercury is placed in the 

 large cavity (d) of the apparatus. The manometer con- 

 taining dry pure nitrogen and terminating in a capillary 

 point is then introduced beneath the mercury, the point 



is broken off, and the bolt inclosing the manometer is 

 screwed into its place. Sections of the steel tubing (c) 

 are then screwed on, one above the other. After the 

 addition of each section mercury is forced into the appa- 

 ratus by the pump, and mounts to the top of the tube ; 

 the height of the column of mercury is measured^ the 

 volume of the compressed gas is read off by means of a 

 cathetometer, and thus the series of observations proceeds 

 slowly until the mouth of the pit is reached. As can 

 easily be imagined, such experiments in the shaft of a 

 coal-pit are by no means easy or pleasant to perform. 

 We can here allude only to the numerous elaborate pre- 

 cautions taken by M. Amagat to insure accuracy in mea- 

 surement and reduce all possible causes of error to a 

 minimum. The divergences in corresponding series of 

 obser\-ation5 never exceeded one-tenth of i per cent. 



Coming now to the results of the experiments made on 

 the compressibility of nitrogen at Verpilleux, we notice, 

 firstly, that the compressibility increases slowly until it 

 reaches a maximum at about 65 atmospheres ; secondly, 

 that it decreases equally slowly until it reaches a normal 

 figure at about 91 atmospheres; and thirdly, that after 

 passing this point it decreases rapidly until at 430 atmo- 

 spheres the volume of compressed gas is five-fourths of 

 what it would be if Mariotte's law were true. In the 

 following table the first column contains the pressures in 

 atmospheres of the column of mercury, the second those 

 deduced according to Mariotte's law from the correspond- 

 ing volumes of compressed nitrogen, and the third the 

 differences between the two : — 



After having established the above table of the changes 

 in the compressibility of nitrogen, M. Amagat was in a 

 position to study the analogous phenomena in the case of 

 other gases with much greater ease. For this purpose it 

 was simply necessary to replace the tube for the column 

 of mercury in the apparatus just described by a mano- 

 meter filled with nitrogen, the counterpart of that used 

 for the gas under examination. By means of these 

 modifications of his original apparatus M. Amagat has 

 prepared very accurate tables for the changes in coinpressi- 

 bility up to 400 atmospheres of air, oxygen, hydrogen, 

 carbonic oxide, ethylene, and marsh gas. In M. Amagat's 

 graphic delineation of the variations from Mariotte's law 

 in the cases of the seven gases mentioned, the abscissae 

 correspond to the pressures in metres of mercury, while 

 the ordinates correspond to the difference between the 

 products of the pressures into the volumes and unity, i.e., 

 to the variations from Jilariotte's law. They all start from 

 a common point— a pressure of 24 metres. The curves 

 of nitrogen and hydrogen are however continued to a 

 minimal measure in accordance with Regnault's data. 

 The minimum ordinate of the ethylene curve, which is 

 425, could not easily be given, 



In glancing over the curves we see that the most 

 conspituous variations occur in the case of those gases 

 most nearly approached to the conditions of liquefaction. 

 The variations in the curve of oxygen are much more 



