74 Notes on a Ghronographic Apparatus. 



Art. VII. — Notes on a Ghronographic Apparatus, with 

 Huyghen's Parabolic Pendulum. 



By K L. J. Ellery, Esq. 



[Read 25th September, 1876.] 



About three years ago, at a meeting of the physical section 

 of this Society, I gave a brief resume of the various methods 

 that had been tried for obtaining uniform rotation, more 

 especially for astronomical and physical instruments ; and I 

 pointed out that as the desired result had been only ap- 

 proached, but in no case obtained, it was a subject worthy of 

 the consideration of the section, and it- consequently formed 

 the matter for discussion at a subsequent meeting. 



It may be as well to state here that all the most success- 

 ful attempts to solve this mechanical problem involved the 

 use of the fly, the rotating or conical pendulum, and reci- 

 procating pendulum, either alone or in combination. 



The governor of a steam-engine is an apparatus the object 

 of which is to secure uniform rotation, and is usually simply 

 a double conical pendulum ; but we know that as the time of 

 rotation of a conical pendulum varies very considerably with 

 the distance the pendulum's bobs are from the axis of rotation, 

 this arrangement alone cannot possibly secure the desired 

 effect, while it usually serves to govern the supply of steam 

 sufiiciently to obtain enough uniformity of motion for the 

 practical purposes of a steam-engine. It is, however to the 

 case of the astronomical or physical chronograph, where 

 absolute uniformity is the most to be desired, and indeed a 

 necessity, that I shall have principally to refer ; and I shall 

 therefore limit my observations to this higher requirement. 



Although the conical pendulum is sometimes used for 

 governing chronographic instruments, it does not, for the 

 reason stated above, afford good results ; if however it were 

 possible to secure a constant driving force and resistance, and 

 therefore a constant arc, it would no doubt be perfect ; but 

 we know it is impossible to attain these conditions. 



In my experiments I have found that a simple free conical 

 pendulum, with a " bob " very heavy in proportion to its 

 length, gives results very near to uniformity if the train be 

 moderately good. 



In order to secure a nearly uniform arc with the conical 

 pendulum many devices have been adopted, most of which 

 depend upon having an excess of driving power and the 

 variable excess used up by friction which is brought into 



