October 1, 1920] 



SCIENCE 



315 



very large compared with the time during 

 which a particular doublet is in the neighbor- 

 hood of B, is proportional to ee' — h' where k' 

 is the mean value of the square of fit) for all 

 the doublets which pass from A to B and arrive 

 at B in this interval. In accordance with our 

 previous hypothesis it seems reasonable to con- 

 clude that Jc' is proportional to the product of 

 the masses of A and B. 



If in the interval of time from i to f + dt, 

 no doublets arrive at B while a doublet left A 

 in the corresponding interval t — (AB/c) to 

 t -\- dt — (AB/c) it is clear that the mean 

 value of the electrical force between A and B 

 in this interval depends on ee' and there is no 

 gravitational action. Other cases may be con- 

 sidered in a similar way and it is clear that 

 the gravitational action depends only on the 

 doublets which go directly from A to B. The 

 action of B on A depends likewise on the doub- 

 lets which go directly from B to A. 



The present theory indicates that there may 

 be a slight screening eilect when a third body 

 C is interposed between two bodies A and B, 

 for C may be supposed to receive some of the 

 doublets which would ordinarily go directly 

 from A to B OT vice-versa. The recent work 

 of ISTipher^ and Majorana^ thus becomes of 

 additional theoretical interest when it is con- 

 sidered in the light of the present theory. 



Gravitational action may be slightly modi- 

 fied, too, by collisions between doublets travel- 

 ling with velocity c. In this connection it may 

 be worth while to point out that if P and Q 

 are two doublets travelling along different 

 straight lines with velocity c, then after a cer- 

 tain instant it is possible for a particle travel- 

 ling with velocity c to meet first one doublet, 

 say P, and then Q hut not for such a particle 

 to meet first Q and then P. A series of mov- 

 ing doublets may thus be arranged in a defi- 

 nite order; something which happens to one 

 doublet may affect those which come later in 

 the series but not those which come earlier. 

 This result may have some connection with 

 the damping of oscillations in the emission of 



2 Science, September 21 (1917). 



^Phil. Mag., T. 39, May (1920), p. 488. 



light. A more imperfect form of the present 

 electrical theory of gravitation has already 

 been published in Proc. London Math. Soc, 

 T. 18 (1919), p. 95, and in the Messenger of 

 Mathematics, T. 48, p. 55. The possibility of 

 a connection with the work of Einstein and 

 Majorana has not been pointed out previously. 

 The present theory seems to be free from the 

 objections raised against the older electrical 

 thory of gravitation (see O. W. Eichardson, 

 "The Electron Theory of Matter," p. 596), 

 there may, however, be some other fatal ob- 

 jections to it. 



H. Bateman 



PROTOZOA IN SAWDUST FOR CLASS WORK 



In studying the method of excreta disposal 

 by compositing night-soil with sawdust, the 

 chance observation was made that microscopic 

 examination of old sawdust piles revealed the 

 presence of Buglypha cysts. Samples of saw- 

 dust were used for experimental culture of 

 hookworm eggs and it was observed that the 

 cultures showed profuse contamination with 

 amoeba, flagellates, ciliates, and free living 

 nematodes. Samples from old sawdust piles 

 were then moistened and incubated with the 

 resuJt that numerous specimens of protozoa 

 and free nematodes were found. 



The sawdust used was chiefly from southern 

 pine. 



This note is published with the thought that 

 it may be of practical service to teachers in 

 providing material for class work. 



0. W. Stiles 

 TJ. S. Public Health Service 



CONCERNING DIASTROPHISM 



Two papers have appeared during the cur- 

 rent year which once again bring before Amer- 

 ican geologists the vexed question of systemic 

 boundaries. In the first Bose^ concludes that 

 the ammonites found at Tularosa, N"ew Mex- 

 ico, 200 feet above the base of the Abo sand- 

 stone, are of CaAoniferous age. This inter- 



1 Bose, E., Am. Jour. Sd., Vol. 49, pp. 51-60, 

 January, 1920. 



