Intelligence and Miscellaneous Articles. 391 



apparatus employed by us in 1863 (for determining the coefficient 

 of dilatation of porcelain), subsequently simplified by the employ- 

 ment of a Sprengel pump, which permits us, as often as we wish 

 it, to take out and measure the thermometric material (nitrogen) 

 contained in the reservoir and to calculate the temperature. 



We have now to describe this method, which is manifestly more 

 expeditious than that (V. Regnault's manometric method) which 

 we formerly employed. 



A cylindrical vessel of Bayeux porcelain, having its two ends 

 spherical, and with a capacity of at least 50 cubic centims., serves 

 as thermometric reservoir. To this a capillary tube is cemented, 

 likewise of porcelain, about 30 centims. in length. This apparatus 

 is cemented to a glass threeway cock, which puts it in communica- 

 tion alternately with the air and, by means of an almost capillary 

 leaden tube, with a Sprengel pump. The thermometer being 

 placed in a partitioned earthen tube with a mixture of asbestus and 

 fireclay, is heated in an oven fed with paraffine oil, the flow of which 

 is regulated by very sensitive cocks. 



The temperature is varied in a very regular manner by a greater 

 or less supply of oil ; it can, at will, be rendered stationary or be 

 raised more than sufficiently to soften or even to fuse the porce- 

 lain*. 



The desired temperature being attained, the flow of the oil is 

 determined so that the nitrogen neither dilates nor contracts any 

 more at that temperature. The fulfilment of this condition is 

 easily ascertained — because the cock carries a drying-apparatus 

 with chloride of calcium, and is terminated by a tube which can 

 be made to dip in the water in order to verify the constancy of the 

 temperature. 



The threeway cock is then turned so as to close all communica- 

 tion between the outer air, the Sprengel pump, and the reservoir. 

 At that instant the leaden tube which joins the threeway cock to 

 the pump is completely emptied; turning the cock again, we 

 exhaust the nitrogen contained in the thermometer. This gas is 

 collected in a graduated tube placed upon the mercury-vat of the 

 pump ; and we only stop when this no longer gives passage to 

 the least bubble of nitrogen. The graduated tube is surrounded 

 with water, the temperature of which is taken. On it is read the 

 volume of the nitrogen aspirated ; the height of the mercury above 

 the level of the vat, as well as the height of the barometer, is de- 

 termined ; and by calculation the volume observed is reduced to 

 that which it would be at 0° and 760 millims. If by a preliminary 

 experiment we have determined the volume of nitrogen contained 

 in the reservoir at 0° and 760 millims., we have the principal ele- 

 ment of the calculation of the temperature attained. 



But it will be noticed that in this apparatus there is a space 

 which is prejudicial ; it is the capacity of the stem of the thermo- 

 meter. 



* This way of heating is so advantageous that we cannot recommend 

 it too highly to those who are occupied with these delicate questions. 



