ON THE DETERMINATION OP GRAVITY AT SEA. 89 



The method of immersing the cistern in a vacuum flask, while satis- 

 factory in preserving the temperature of the enclosed air at a nearly constant 

 value during each run, involves a difference in temperature between various 

 parts of the barometer tube, and this gradient cannot accurately be allowed 

 for. On the homeward voyage, when temperature conditions were noto- 

 riously adverse, there was sometimes a difEerence between reservoir and 

 stem which amounted to 3° or 4°. Even on s.s. Ascanius, where conditions 

 were exceptionally favourable, there was seldom a difference less than 

 0°'2 to 0°"3. These considerations involve errors much larger than the 

 differences in gravity which it is sought to discover. 



The aneroid method with a plain mercury barometer is in this respect 

 better than one which involves a gradient within the tube itself, though 

 it has other objections. 



It is the wi-iter"s opinion that all barometers used for gravity work 

 must be of uniform temperature and that they should be immersed in a 

 well-stirred bath of liquid which is kept at as constant a temperature as 

 possible by a thermostatic device. 



It is the writer's opinion that Hecker's method of an enclosed barometer, 

 which is photographically recording, could, if modified in the respect 

 indicated, be made to yield satisfactory results at sea. 



In concluding this report upon the three methods tested dm^ing the 

 voyages to and from Australia of the British Association in 1914, the writer 

 expresses his thanks to Professor Hecker for giving him the opportunity 

 of testing the apparatus and to the Council of the British Association for 

 their grant from the Caird Fund for the purpose of these tests, which are 

 only to be regarded as prehminary to what it is hoped may prove a suc- 

 cessfid attempt at a later date. For this purpose apparatus is already 

 in course of preparation. 



II. The Influence upon a Marine Barometer op the Ship's Motion 

 THROUGH the Air. By Professor W. G. Duffield. 



Before a further attempt is made to determine the value of gravity at 

 sea there are certain jKoblems to be solved. These are chiefly concerned 

 with the behaviour on board ship of a mercury barometer, which in one 

 form or another is employed in all methods to which extended trials have 

 been given. 



Both Hecker and the writer have had reason to question the accuracy 

 of a comparison between readings made in harbour and those on the high 

 seas, and a far more careful examination of the effect of the ship's vibration, 

 due to the throbbing of the engines, as well as to the tossing on the waves, 

 is necessary. 



It was with the object of testing this point amongst others that the 

 writer sought and obtained permission to carry out experiments on one 

 of His Majesty's Destroyers, and in August 1919 two marine barometers 

 were mounted on board H.M.S. Plucky, one in the chart-room below the 

 bridge, where the vibration appeared to be least, and the other in the ward- 

 room, where it was greatest. It was intended to compare their readings 

 with one another and with those of a barometer on board a stationary 



