272 SCIENCE PROGRESS 



ASTRONOMY. By H. Spencer Jones, M.A., B.Sc, Royal Observatory, 

 Greenwich. 



The Gravitational Constant and Temperature. — An important 

 memoir by Dr. P. E. Shaw has recently appeared {Phil. Trans. 

 R.S., ser. A, vol. ccxvi. pp. 349-92, 1916) in which, as a result 

 of an extensive series of experiments, the author announces 

 a dependence of the Newtonian constant of gravitation upon 

 temperature. The experiments lead to the conclusion that 

 " when one large mass attracts a small one, the gravitative 

 force between them increases by about 1/500 as the temperature 

 of the large mass rises from, say, 15 C. to 215 C." It is not 

 possible to say definitely that the relation is a linear one, but 

 assuming that this is so the result may provisionally be stated 

 in the form that the constant of gravitation has a temperature 

 coefficient of + 1*2 x io -5 per i°C, under the conditions just 

 specified. 



As far as can be judged from the account given in the 

 paper, the experiments have been carried out with great care 

 and thoroughness and precautions seem to have been taken 

 to eliminate every possible source of error. Whether or not 

 the startling result which Dr. Shaw has announced will be 

 substantiated by later experiments, a tribute should be paid 

 to the care with which he has carried out these difficult and 

 delicate experiments. The method used was an elaboration of 

 the original Cavendish experiment — consisting in measuring 

 the differential attraction of two large masses on two small 

 masses when the position of the large masses is changed. The 

 small masses were suspended by delicate quartz fibres in a tube 

 from which the air could be pumped so as to leave a very high 

 vacuum. Special precautions were taken to avoid any possible 

 errors which might arise from electrostatic or magnetic actions, 

 and the pressure of the residual gas left in the exhausted tube 

 was such that the effects of radiation pressure and convection 

 were a minimum. The pressures used were about 14 mm. 

 and 20 x io -6 mm. The former pressure has the advantage 

 that the residual air is sufficient to damp the torsion system 

 in an efficient manner ; to overcome the trouble arising from 

 the lack of damping at the very low pressure, the ingenious plan 

 was adopted of making the suspended weights in the form of 

 chains, so that the vibrations might be damped out by the 



