158 Intelligence and Miscellaneous Articles. 



Thus in both groups the duration increases as the pressure 

 diminishes. 



To determine the magnitude of this variation between two deter- 

 minate pressures, it is enough to divide the increase of the duration 

 by the total duration at the higher pressure, and by the difference 

 of the pressures. A coefficient is thus obtained which expresses the 

 mean increase of the unit of duration (I s ) for a diminution of 1 milli- 

 metre in the pressure. The differences in the fuses of the first 

 group were too great to be safely used in calculating this coefficient ; 

 those of the second group give the following values : — ; 



Between Ouchy and St. Pierre 0*00104 



Between St. Pierre and Chenallettes 0*00108 



These two coefficients, which are almost identical, show that the 

 increase in the duration of the combustion is proportional to the 

 diminution in pressure. This very simple law has been already 

 enunciated by Prof. Frankland. 



To compare the values I have obtained with those of Messrs. 

 Mitchell and Frankland, it is enough to calculate by the aid of 

 their results the millimetric coefficient between limits of pressure 

 nearest those under which I operated. Taking Mr. Mitchell's ob- 

 servations in the Himalayas at pressures of 752 and 584 millims., the 

 coefficient is 0*00161 ; between the pressures 752 and 609 millims. 

 it is 0*00140. Thus the increase was a little greater in these fuses. 

 Of the six pressures in Prof. Frankland's experiments, the two 

 which approach most closely to the limits within which I worked 

 are the^second, 716*8 millims., and the fourth, 570*2 millims. He 

 found for the duration of the combustion : 



millims. s. 



A 716*8 32*25 



A 570*2 37*75 



from which is deduced the coefficient of variation 0*001 1 6 



Between Ouchy and Chenallettes my experiments give. . . . 0*00111 

 This is certainly a remarkable agreement, and the more interesting 

 as Prof. Frankland's fuses differ both in dimensions, shape, and 

 duration from those which gave the above results. Hence it may be 

 admitted that the duration of the combustion of one of these fuses 

 increases on the average by 0*0011 of its value for each diminution 

 of a millimetre on the pressure. 



These facts have an importance which cannot be neglected in a 

 military point of view, and they ought to be taken into account 

 whenever the duration of the combustion of a fuse is an essential 

 element in its use. 



As to the cause of this increase of the duration of combustion 

 when the density of the surrounding air is less, one would at first 

 sight be tempted to ascribe it to a diminution of the oxygen. This 

 idea, however, is not tenable, as the fuses contain enough burning 

 substances in the form of nitrate. To convince myself directly that 

 atmospheric oxygen does not intervene, I burned three fuses in a large 



