R. Bunsen — Calorimetric Investigations. 281 



is now for the most part removed with a pipette from the tube 

 c, the tube is freed from all moisture by means of a current of 

 diy air produced by the water air-pump, and only then is the 

 iron head-piece d cemented on with the finest sealing wax, in 

 such a manner, that the tube c projects inside to some distance 

 above the bottom, in order that later, the stopper bearing the 

 scale may rest in the mouth of the glass tube c and not in the 

 iron casing. The last filling with boiled mercury to the level 

 ;- is performed by the aid of a capillary glass tube, in order 

 to avoid all air bubbles on the sides of the tube. 



In order to place the mercury thread in each experiment on 

 one of the first divisions of the scale, it suffices to press the 

 cork of the graduated tube, with a rotary movement, somewhat 

 deeper into the mercury tube c, fig. 1. If the thread has there- 

 by exceeded the commencement of the scale, there is a small 

 brass weight, fastened on a thread and previously warmed by 

 contact with the hand or tongue, sunk into the fluid a, fig. 1. 

 If the weight of the brass weight be g gTams, its temperature 

 U Its specific heat s^, the latent heat of melting for water /, and 

 the weight of melted ice, obtained by equation (2), which cor- 

 responds to an oscillation of one scale division, denoted by 

 P, then the oscOlation produced by the brass weight g is equal 

 to ^~ scale divisions. If we let 



L grm, 5 scale divisions. I 0*6 grm. 30 scale divisions. 



By dipping in one of these brass weights, previously warmed 

 uncier the tongue, the mercury thread can be made to retreat to 

 tlie desired extent. The oscillations corresponding to the Httle 

 J^eights, which were only warmed to 37° C, are well adapted 

 ^give an idea of the extraordinary deUcacy of the instrument. 

 ^*^^^, Increase of temperature which 04 grams of brass at 

 07 C. would produce by immersion in the mass of water in the 

 ^strument, amounting to about 20 grams would displace the 

 mercury thread of a centigrade thermometer only 0° "07, that 

 ^i the just described calorimeter, however, twenty scale divi- 

 ^?ns, each of which was, in the instrument employed, one 

 »»iUimeter in length. 



tuf ri^tion to the readings on the scale, it is still to be observed 

 \^^^. before such observation of the scale tube, which is best to 

 . e made with the telescope, the former must, particularly when 



's very narrow, be gently agitated by repeated rapping until 



