928 



SCIENCE. 



[N. S. Vol. XIX. No. 494. 



by a particular magma depends chiefly on the 

 pressure to which it is subject. (The in- 

 fluence of temperature is not important in the 

 present connection.) As pressure is gradually 

 relieved during eruption, more and more of 

 the contained gas is discharged. When ex- 

 plosion of vesicles is once initiated at the 

 top of the column it reduces the pressure on 

 lower parts by carrying away some of the lava, 

 and this loss of pressure in turn promotes the 

 exclusion of the gas. If this view of the 

 process is correct, the paroxysmal character 

 of explosive volcanic eruption is strictly an- 

 alogous to that of geyser eruption. 



These theoretic considerations lead to the 

 prediction that when the Pelee spine shall 

 have become so cool as to permit of close in- 

 spection, its lava will be found to be porous. 

 Porosity may, perhaps, not be accounted a 

 verification of the theory, but the absence of 

 vesicles would prove it untenable. 



G. K. Gilbert. 



Washington, 

 May 26, 1904. 



A SUGGESTIVE RELATION BETWEEN THE GRAVITA- 

 TIONAL CONSTANT AND THE CONSTANTS OP 

 THE ETHER. 



The phenomena of radio-activity and the 

 ionization of gases point so strongly toward 

 the electrical constitution of matter, that the 

 writer has made an attempt to connect the 

 fundamental constant of gravitation with the 

 electrical constants of the ether. 



The result obtained is published with the 

 hope that it may suggest to other physicists 

 a more valuable extension. 



The gravitational equation as ordinarily 

 written is 



where k is the gravitational constant and 

 M, IT are the gravitating masses. The unit 

 of mass is the gram. This is a purely arbi- 

 trary unit, so I have chosen a new unit of 

 mass, which may be defined as follows : The 

 unit of mass shall be that mass which is asso- 

 ciated with one electromagnetic unit each of 

 positive and negative electricity. This mass 



is considered to be made up of electrons, each 

 of which has a definite mass associated with 

 a definite amount of electricity. 



The adoption of this unit of mass involves 

 a change in the numerical value of the gravi- 

 tational constant. The object of this paper 

 13 to investigate the value of this constant. 



The ratio of the charge to the mass of an 

 electron as well as the charge itself has been 

 determined by direct experiment. The most 

 probable value of the charge e on an electron 

 is 10—20 electromagnetic units, as measured 

 by Mr. H. A. Wilson. The ratio e/m has 

 been measured by a number of physicists. 

 The following are some of the best values 

 found by experiment: 



Kaufmann (1898) 1.86 x 10' 



Simon (1899) 1.865x10' 



Lenard (1899) 1.15 s 10' 



Kaufmann (1901) 1.31 x 10' 



Wiechert (1899) 1.42 xlO' 



The mean of the above values is e/m ^ 1.52 

 X 10' electromagnetic units. This quantity as 

 well as the charge e is probably correct to one 

 significant figure. 



The charge e as stated above is 10 — 20 . bence 

 the mass of an electron is m ^ .65 X 10 — ^^ 

 grams. 



The number of electrons carrying one elec- 

 tro-magnetic unit of electricity is 10"*; con- 

 sequently the mass associated with one unit of 

 negative electricity is .65 X 10 — ^ grams. 

 Gravitating matter as we know it is neutral as 

 regards charge. There must be present then 

 an equal amount of positive electricity. The 

 mass associated with this positive electricity 

 will also be .65 X 10 — "^ grams ; hence the total 

 mass that is associated in the combination of 

 one unit each of both electricities is 1.3 X 10—'^ 

 gTams. 



This is the new unit of mass. The new 

 gravitational constant may be found by sub- 

 stituting in the equation 



,.2 



The value of h for gram unit of mass is 

 6,673X10—", from which 



1 ^ 1 ^ J_ 



9 X lO^V ~" (3 X 101°) V hh-^' 



