354 



Mr. J. Joljc 



ference obtaining in all other methods is almost entirely avoided, and 

 in fact the facilities offered by the method are more fully availed of. 

 We have in fact a change of state method which involves, as I hope to 

 show, no sensible radiation errors, the simplest thermometry and appa- 

 ratus, and no manipulative difficulties whatsoever. 



Plate 6 is from a photograph of the apparatus. The calorimeter 

 consists of a nearly spherical chamber of very thin spun brass, 

 10*5 cm. in diameter, made up of two hemispheres meeting on thin 

 ground flanges. These hemispheres are carried by standards which 

 slide upon the centre one of three rails forming a light girder. This 

 girder is supported on three feet, and is arranged to stand in one 

 assigned position inside the case of a large physical balance. I used it 

 for a considerable time exactly in the position of the bridge used for sup- 

 porting a beaker in taking specific gravities ; that is, spanning one of 

 the pans and passing between its stirrups. I subsequently found it 

 necessary to remove the pan, substituting a counterpoise. It will be 

 understood now that when the standards carrying the hemispheres 

 are slid along the centre rail, the sphere can be completed inside the 

 balance case, and when meeting about the centre of the rail, may be 

 made to enclose a substance depending from the hook of the balance. 

 To permit of the substance being weighed while thus enclosed in the 

 sphere, a notch is provided in the flanges at their top meeting point. 

 This notch is, for reasons to be explained later, cut in two little plugs 

 of plaster of Paris, contained in cells provided for them. The sphere 

 has two openings with short tubulures, one on each hemisphere, 

 horizontal, and opposite one another. The larger of the two, 2*2 cm. 

 diameter, opening to the front of the balance, serves for the admis- 

 sion of steam; the other, which when in use is prolonged by the 

 addition of a couple of centimetres of rubber tubing, is about 1 cm. in 

 diameter, and serves for the exit of steam at the rear of the balance. 



To secure the more sudden admission of steam, a movable admis- 

 sion pipe fitting the front nozzle on a ground joint is provided. It is 

 carried on a separate standard, also fitting the centre rail, and is con- 

 nected with a small tin or copper boiler* by a flexible tube, so that 

 when it and the rubber tube connexion have been first thoroughly 

 heated by the passage of steam, it can be run up on the centre rail to 

 meet the nozzle and discharge the steam directly into the sphere. 

 Over the entire length of the calorimeter extends a rectangular 

 tunnel-shaped covering of mahogany, made in halves, meeting at the 

 centre of the girder, and provided with notches at their meeting faces, 

 also cut in plaster of Paris, to permit the passage of the wire sup- 



* I use a cylindrical copper boiler, 15 cm. diameter by 20 cm. higb. Tbis, beated 

 by a strong Bunsen burner, evolves steam at tbe rate, I find, of about *4 gram 

 per second. As it is important to introduce tbe steam as quickly as possible, a 

 large boiler is advantageous. 



