216 MESSRS. R. THEELFALL AND J. A. POLLOCK 



ingenuity which have hitherto been submitted, must necessarily have failed from the 

 difficulty of securing in them the requisite uniformity of temperature. 



With regard to forms of apparatus in which advantage is taken of the elastic 

 properties of gases, it is proper to observe that, though it is possible to arrange a 

 compensation effective to the first order of the variations due to change of tempera- 

 ture, yet the apparatus must almost necessarily be too large to satisfy the condition 

 as to the temperature being sufficiently uniform throughout. 



Very closely associated with the problem above discussed is that of making an 

 instrument by which small secular variations in the gravitational force at any one 

 place may be ascertained, though in this case the problem is very considerably 

 simplified, because the question of portability does not arise, and arrangements may 

 be made for minimising fluctuations of temperature. In this case, however, the 

 observations should be of a higher order of accuracy than is necessary during a gravity 

 survey. 



As soon as we became familiar with the properties of fused quartz, it became obvious 

 to us, as well as to others, that by taking advantage of these qualities it might be 

 possible to construct a balance having sufficiently permanent elastic properties to give 

 a practical solution of the problem of constructing a gravity meter. 



A committee of the British Association, which in 1886 had invited designs for a 

 gravity meter, reported in 1889 that work had been suspended pending a trial of 

 fused quartz. Our own attempts to construct a gravity balance began in September, 

 1889, and have continued uninterruptedly ever since. 



When we began to work at the matter we formed an impression that the problem 

 of observing small variations in the intensity of gravity at any one place would prove 

 simpler than that of constructing a portable instrument, and consequently we first 

 turned our attention in this direction. We worked at instruments of what may be 

 called the non-portable balance class for two years, at the end of which time we had 

 satisfied ourselves that we were not likely to attain to sufficient sensitiveness, and 

 accordingly we turned our attention to the construction of a portable instrument. 

 We experimented in this direction for some time, with the result that in October, 

 1892, we began to construct what we hoped would prove to be a final form of instru- 

 ment, but it was not until September, 1893, that we had got it forward enough to 

 commence systematic observing. 



With regard to portability we may say that we have travelled with the present 

 balance over 6,000 miles. Of seven quartz threads which we have had in actual use 

 since the balance was completed in September, 1893, only one has broken. During 

 the first journey in February, 1894, and through the negligence of a person who had 

 undertaken to look after it, the balance was knocked off its stand and practically 

 destroyed; the only part which was not broken was the thread. The present tliroid 

 was mounted on the 10th of September, 1896. It has travelled over 4,600 miles by 

 cart, railway, and steamer. Observations have been taken with this thread at four 



