On Negative Fluid Pressure on a given Surface. 207 



solved, and, after standing, rapidly cooled, gave a.bromate of 

 potassium, which was washed out with cold water, thrice crystal- 

 lized, and then fused. For 100 parts of silver, 110*357 parts 

 of this bromide of potassium were obtained. The same salt was 

 dissolved in bromine-water, evaporated, the residue heated to 

 about 400°, again dissolved, and the solution, which was not 

 perfectly clear, filtered through spongy platinum and evaporated 

 to crystallization. Of this salt, there were 110'334 parts for 100 

 parts of silver. The bromide of potassium in the residue from 

 this salt exhibited the ratio 100 Ag : 110-335 parts K Br. 



As the general mean, 100 parts Ag correspond to 110345 

 parts K Br. Marignac found that 100 parts correspond to 

 110*343 parts K Br. We give in conclusion the average 

 numbers for the atomic weights determined by the author : — 



9 = 16. H=l. 



. =107-930 107*660 



14-044 14-009 



79-952 79-750 



35-457 35-368 



. 126-850 126-533 



7-022 7-001 



39-137 39-040 



23-043 22-980 



16-000 15-960 



Silver 

 Nitrogen 

 Bromine 

 Chlorine 

 Iodine 

 Lithium . 

 Potassium 

 Sodium . 

 Oxygen . 



XXVI. On Negative Fluid Pressure on a given Surface, 

 By Charles Brooke, M.A., F.R.S. 



To the Editors of the Philosophical Magazine and Journal. 

 Gentlemen, 



THE perusal of two papers in the last Number of your valu- 

 able Journal, containing some highly interesting experi- 

 ments on fluid motion, but in which I have sought in vain for 

 any definite exposition of their physical causes, has induced me 

 to send you a brief (unpublished) paragraph from my chapter 

 on Hydrodynamics, which will, I think, help to place the origin 

 of these and other analogous phenomena in a somewhat clearer 

 light. 



In the well-known experiment of Bernouilli, in which a current 

 of fluid passing through a diverging conical tube draws in addi- 

 tional fluid through a lateral branch, and also in that of De- 

 sormes, in which a stream of air issuing from an orifice in a 

 plane surface virtually attracts a circular disk placed parallel to 

 and in front of it (which I have found to hold good equally with 

 a jet of water), and in other similar phenomena, the cause has 



