]^^ ACrrON OF HEAT MODliILD 



"dnctly rounded alr-holcs. The fragments of shell which haiJ 

 occupied the upper part of the little tube, had lost every trace^ 

 of their original shape in the act of ebullition and fusion. 

 The compress- No. 7-— On ttie '26th a similar -experiment was made, ia 

 .eTwith^explo-* ^^^^J^h the baiTcl was thrown open, in spite of this powejful 

 t^ion- compressing for®c, with a report like that of a gun, (as 1 was 



told, not hav-ing been present), and the bar was found in a state 

 of strong vibration. The carbonate was calcined, and some- 

 what frothy, the heart of one piece of chalk used was in a state 

 of saline marble. 

 Operations It now occurred to me to work witb a compressing force, 



^I*rT '^•' V t>' ^"^ "" air-tube, trusting, as happened accidentally in one case, 

 that the expansion of the liquid would clear itself by gentle 

 exudation, witliout injury to the carbonate. In this mode, it 

 was necessary, for reasons lately stated, to place th^muzzle 

 upwards. Various trials made thus, at this time,*'afforded no 

 remarkable results. But I resumed the method, witli the fol- 

 lowing alteration in the application of the weight, on the 27th 

 of April 1S04.. 

 A direct com- I conceived that some inconvenience might f^risc from the 

 ii&eil and the i^^o^le of employing the weight in the former experiments. In 

 lever njected. them it had been applied at the end of the bar, j^nd its effect 

 propagated along it, so as to press against thebiirrel at its other 

 extremity. It occurred to ujo, that the propagation of motion 

 in this way, requiring some seuiiblc time, a considerable quan- 

 tity of carbonic acid might escape by a sudderj eruption, before 

 that propagation had taken elject. I therefore thought, that 

 - more citectui^l work might be done, by placing a heavy mass, 



(fig. 40.), so as to act directly and sijuply upon the muzzle of 

 the barrel ; this mass being guided and co;nniandcd by means 

 of a powerful lever, fa bj. For this purpose, I procure an 

 iron roller, weighing 3 cwt, 7lh/., and suspended it over the 

 furnace, to the end of a beam of wood, resting on a support 

 near the furnace, with a long arm guided by a rope Cc c) and 

 pulley (d), by which the weight could be raised or Jet down 

 at pleasure. 



With this apparatus I inadc sopiic tolerable experiments ; 

 but I found the weight too light to afford certain and steady 

 results of the best quality. I therefore procured at the foundry 

 a large mass of iron (J J, intended, I ticlieve, for driving piles, 



and 



