42 DENSITIKS or (IXYCKN AND IIM)R()(iKN 



Most of the objects weighed with this a[ti);iiatus have to be couliiiued iu 

 desiccators during the weighing. In order to carry these desiccators around with 

 the objects which they contain, it is rei^uired to let the phitfonn /i" E, which is 

 placed eleven decimetres below //, rotate with the latter. E E '\h supported on a 

 pivot fixed in the floor of the closet, directly under the centre of//. To the under- 

 side of the disk s- is fastened tli.- pl.mk C <", and the latter is joined to E Ehy two 

 boards seen in section at J) D. 



When it is convenient to place two pairs of globes on the api)aratus at once, 

 a desiccatoi- containing the first pair stands on this platform E 7i, and the globes 

 are suspended by long wires to the hooks near ee in Fig. 19 (this arm should 

 have been shown parallel to the k)nger sides of E E). Across this desiccator, 

 is then placed another, whose globes are suspended from the hooks seen near the 

 latches x x. The central pai-t of this upper desiccator has to be somewhat con- 

 tracted so as to leave free the wires suspending the lower globes. There is then 

 still room for a tliii'd pair of objects of no great bulk, like the tube containing 

 palladium which is to be mentioned in describing the experiments on hydrogen. 

 Since it was commonly thought proper to leave objects on the balance as much as 

 twenty-four hours, these facilities for weighing three objects within the interval 

 required for one were a very great convenience. 



The desiccators for containinsr the ijlobes needed to have dry air bi'ouglit to 

 tiiem from without. In order to maintain this current while the globes were in the 

 balance closet, it was convenient to introduce it thiough the axis of the reversal 

 ap[)aratus. For this purpose, the stud on which the api)aratus i-otates was made 

 hollow, as seen at the centi-e of the plan, and the liose bringing dry air was attached 

 to the upper end of this tube. It ended lieneath in a little clianibcr, turning 

 with the revolving object carrier, to which could be connected rubber tubes 

 leading to the desiccators. The air was forced by a hydraulic blower through 

 two carboys each half full of pure sulphuric acid. After bubbling through the 

 acid, the air was filteied through a long column of cotton-wool. 



Even when the mechanism w.-is turned slowly, it was not easy to avoid some 

 oscillation of the objects carried on the reversal apparatus. A mechanism foi' stop- 

 ping this oscillation was therefore devised. The hooks a a are suspended from the 

 pans of the balance by rods which are enclosed in the l)rass tubes on which the 

 pans of the Kueprecht balance are su})ported when the lialance is arie.sted. On 

 these rod.s, two brass disks slide IVccly down to two j.iiis wliicli Jiinlt the motion. 

 In the ttiijes are placed perforated disks coveicd with velvet. W hen llie tubes 

 carrying the disks and the pan arrests are let<lownsix millimetres, ihe sliding disks 

 still rest on the velvet. The iians have at this moiniMit been fi-ee(l fi-oni their ar- 



