232 MESSRS. R. THRELFALL AND J. A. POLLOCK 



passes through a stuffing box and fits on the end of the arrester shaft. The milled 

 head may be clamped by a clamp screw on the end of the balance case, and thus we 

 can make sure that the arrester does not work loose during transport. The thread 

 being quartz in a perfectly dry atmosphere will of course insulate perfectly. We have 

 a check on the possible electrification of the lever because we can set it swinging so as 

 to touch the open jaws, and we can then observe whether any change in the reading 

 occurs. We have often suspected that we were troubled by electrostatic effects but 

 we have no conclusive evidence that this was really the case. 



The whole of the inside of the balance is blacked so as to increase the surface con- 

 ductivity, and so enable the temperature to equalise itself with the greatest possible 

 facility. 



In fitting the instrument together the surfaces are coated with soft wax before 

 being screwed up. This makes all the connections air-tight. The last operation 

 after putting the machine together is to exhaust it down to about 24 inches of 

 mercury pressure, and then admit air dried by sulphuric acid and phosphorus 

 pentoxide. The air is also passed over potash and through a filter of cotton wool, the 

 object being to have dry dust-free air in the balance case. 



The process of exhaustion and readmission of air is repeated many times ; in 

 making a new instrument we would have two openings into the balance and draw dry 

 air through it instead of exhausting and readmitting air. Finally, the balance is 

 left with air at a pressure of about 25 inches of mercury, but, of course, the 

 actual pressure selected depends on the temperature of the balance at the time ; a 

 temperature of about 25 C. would be suitable for the pressure named. It is, of 

 course, our object to have such a degree of exhaustion that -there will always be an 

 inward pressure wherever the balance is taken, and yet not to have a greater pressure 

 difference than is necessary. A survey in mountainous regions would naturally 

 require a higher degree of exhaustion. 



Mounting the Riding. Level. 



The mounting of a sensitive bubble tube, such as is employed in our riding level, is 

 a matter calling for attention. We have availed ourselves of the experience of our 

 friend, Mr. G. KNIBBS, and, acting on his advice, have mounted the glass tube in a 

 stout tube of copper, and packed it in position by means of glass wool. The packing 

 is tight enough to put any displacement of the bubble tube out of the question, and yet 

 the tubes are free to follow their own tendencies in the matter of expansion or con- 

 traction. We have proved by experience that a level mounted on the tube of the 

 balance case is not sufficiently to be relied upon, and this at least suggests that some 

 of the irregularities we have observed may be due to warping of the metal work 

 under the influence of changes of temperature. In the event of such warping 

 exceeding the elastic limits of the material, we should have permanent sets which 



