187 



distributed between the molecules which the a particle 

 traverses. There does not appear to be any evidence, as yet, 

 that the chance of an ion being formed from a molecule is 

 dependent on whether the molecule has already lost one or 

 more ions; rather the contrary. If this is the case, occa- 

 sional molecules must lose several ions. Nor is it yet clear 

 in what mode ionisation occurs. Does the a particle simply 

 cause the removal from the molecule of one or more electrons ? 

 May there not possibly be a more complete disruption of the 

 molecule, or even the atom ? There is one curious parallelism 

 in numbers which may have a bearing on this question. Ram- 

 say and Soddy (Proc. Roy. Soc, 72, p. 204, 1903) found 

 that 50 mmg. of radium bromide in solution evolved gases 

 at the rate of '5 cc. per day — /./., 2 x 10'" molecules per day. 

 Now, Rutherford has shown that one gram of radium bro- 

 mide, without its radio-active descendants, produces 3'6 x 

 10'" a particles per second. Each a particle makes 86,000 

 pairs of ions. Hence the number of ions made in one day 

 by 50 mmg. is 



3-0 X 10"° X -05 X 60 X 60 x 24 x 172,000 = 27 x 10^-'. 



This number is an inferior limit. A superior limit is 

 found by considering all the radio-active products of radium 

 to be present in full, in which case the number will be be- 

 tween five and six times greater. The close agreement of 

 these numbers certainly fits in with the hypothesis that an 

 actual disruption of the w^ater molecule takes place in con- 

 sequence of the passage of the a particle through it. 



I owe my thanks to my assistant, Mr. A. L. Rogers, for 

 the great care and skill with which he has made the apparatus 

 used in this work, and drawn the plate illustrating this paper. 



Note : October 22. — The greater part of § 3 of this paper 

 ha been written since the remainder was read. 



DESCRIPTION or PLATE YII. 



Insulators shown by dotted surfaces. 



QQ, upper plate of ionisation chamber. 



P, one of three glass pillars supporting gg. 



gg, upper gauze, forming lower plate of iouiisation chamber, 

 connected to battery. 



g'g', lower gauze, supported by three metal pillars, one of 

 which is shown, earthed through pillars and metal of case. 



SS, semaphore of thin sheet metal, worked from without by 

 turning the long rod on which it is mounted. 



TT, set of vertical tubes. 



RR. radium plate. 



DEFG, outline of electric oven. 



