ON THE DETERMINATION OF GRAVITY AT SEA. 87 



though had conditions been more favourable they would have been taken 

 into account. 



From the dimensions of the apparatus and the coefficients of expansion 

 of its various parts it is possible to calculate the relationshiji between 8g 

 and the variations in T, the temperature of the air reservoir, in t, the stem 

 temperature, and in h, the height of the mercury column in millimetres. 

 But, in view of the difficulty of accurate determinations of the volumes 

 of the various parts of the barometers (though this was tried), it 

 was considered more convenient to use observations made in harbour 

 stations, where gravity was known from pendulum observations, for the 

 determinations of the coefficients in the equation 



83 = AJh + B,ST + CJt. 



From a knowledge of A^, B^, C^ and the variations in h, T and t from 

 their values at a primary standard station, the variation, 8j, from the 

 value at that station is theoretically obtainable for any observation made 

 during the voyage. 



The Results of the Test. 



The generous provision by Messrs. Holt of a laboratory in the refri- 

 gerator of s.s. Ascanitis has been referred to in previous reports and the 

 excellent results as regards temperature regulation put on record. It 

 was with the utmost disappointment that between Cape Town and Austra- 

 lia after nearly four weeks of continuous observation, it was discovered 

 that all four barometers, instead of being sealed up, had developed leaks, 

 and that they were resj)oncling to the changes in the external atmosjDheric 

 pressure. This was in spite of great care on the part of the writer in 

 assembling the apparatus, the whole reservoir having been coated, after 

 screwing up, with white lead paint. It is believed that the leak occurred 

 where a glass tube and tap runs into the cistern, and that a much better 

 fit would be possible. 



By collecting all the paraffin candles on board and melting them in a 

 pot, into which the barometer tubes were immersed to about 8 inches, the 

 writer, before reaching Fremantle, had sealed up all the barometers. But 

 the opportunity of using the excellent refrigerator laboratory for the 

 return voyage vanished on arrival in Adelaide, when it was found that 

 war had broken out and that the ship had been commandeered as a 

 troopship. 



It has already been exjjlained in the Interim Report that accommodation 

 was found on R.M.S. Morea for the return voyage, but that the tempera- 

 ture regulation was exceedingly inefficient. Nevertheless, an attempt 

 was made, and after resealing the barometers in Sydney Harbour, the 

 whole of the foregoing operations were carried out on the Morea from 

 September 9 to October 20, 1914. 



It was not until the Armistice that the writer found an opportunity 

 of attempting the immense labour of the measurement of the films and 

 of the reduction of the results, but that occasion found him vnth a staff of 

 officers of the Royal Air Force, some of them with considerable qualifica- 

 tions for the work. Other work being in abeyance, they readily agreed to 



L 2 



