216 Proceedings of the Royal Society of Victoria. 



now. On these grounds I should view with suspicion any design 

 which included a flexible bar ; among these may be mentioned 

 the pendulums lately employed by Professor Mendenhall,* which 

 were of the same general pattern as the Indian pendulums, 

 though much shorter, and consequently less liable to injury. The 

 best pattern I have seen is that adopted by Lieut.-Col. von 

 Sterneck, which consists of a rigid rod, carrying at its lower end 

 a bob composed of two frustra of cones united at the base. 



Recent practice inclines more and more, and rightly so, to the 

 use of half-second pendulums ; for a pendulum of one-fourth 

 the length is, at least, ten times as free from risk of injury in 

 handling. A five-foot bar is distinctly awkward, but a thirteen- 

 inch bar is just about as comfortable a size to handle as could 

 well be imagined, to say nothing of the increased rigidity. This 

 method secures the further ad vantage of portability, a matter of 

 great importance in survey work. 



Not. merely should the pendulum be short and rigid, its 

 structure should be, as far as possible, continuous throughout. 

 This, unfortunately, cannot be fully attained; the best which 

 can be done is to have the different parts so made that 

 they can be attached together by a process of shrinking on 

 and riveting. Solder is always objectionable, and its use can be 

 avoided by employing a suitable design. 



One great obstacle to the attainment of invariability is tlie old 

 practice of attaching the knife-edges to the pendulum and 

 swinging it on a plane suspension ; for the knife-edges get dulled 

 and have to be reground ; each regrinding of course changes the 

 effective length of the pendulum, and destroys the differentiality 

 of the observations. Professor Mendenhallf took a great step in 

 advance when he attached the planes to the pendulum, and 

 swung it on a knife-edge suspension : for a good agate plane 

 requires only to be kept clean ; while the knife-edges, if separate 

 from the pendulum, maybe reground at pleasure without sacrifice 

 of differentiality. Incidentally this mode of suspension introduces 

 a good many other improvements, which may as well be summa- 

 rised here. 



» U.S. Coast and Geoaetic Survey Report for 1891, appendix 15, p. 503. 

 t L.C.. p. 5.30. 



