1869.] The Lamheth Ohservatory. 345 



a firm foundation upon so unpromising a base as a river-bank 

 shaken by the incessant rolUng of heavy carts, has been found to 

 be very efficient, if not practically perfect. Since the first setting 

 of the masomy of the platform not the slighest permanent change 

 has occurred in the bearings of the instruments. Passing carts 

 communicate a slight transient vibration for the moment, which, 

 however, has no other inconvenience than causing an occasional 

 brief suspension of observation. 



Upon entering this little model testing observatory of the 

 Indian store, then, the general aspect is that of a square room with 

 a raised circular platform in its centre, reached by a small flight of 

 stairs. A broad solid wall-like rim, breast high, and included 

 within an outer suspended pathway for the observer, runs round 

 the platform, and aflbrds a convenient resting-place for instrumental 

 appliances of various kinds. In the midst of this circle a flat-topped 

 pillar forms the bed upon which the instrument to be tried is placed 

 by the examiner. 



Overhead the broad glass roof, with the uninterrupted space of 

 clear sky, indicates that the star-rulers of the night can be appealed 

 to whenever such higher and more refined arbitration is deemed 

 desirable. 



One of the most important objects that is attempted in this 

 observatory is the examination of the exactness of graduated circles 

 that are designed for horizontal measurement. Unless these circles 

 do measure equal and true degrees in all their parts, it is obvious 

 that the angular intervals recorded from them are not worth the 

 paper upon which the records are inscribed. To accomplish this 

 end four horizontal tubes, called collimators, have been placed on 

 difierent parts of the circular wall, so that an observer can contem- 

 plate each of them from the telescope attached to the instrument 

 under trial. These colhmators are in reality only so many fixed 

 and immovable points, occupying for the time known, or more 

 properly ascertainable, positions on the great wall-circle, and there- 

 fore including also known, or ascertainable, angular intervals 

 between them. The exact reference-j)oints, or virtual centres in 

 these collimator tubes — " the marks to be collimated or aimed at " 

 — are of a varied character. In one there is a system of diagonally 

 crossing spider-threads, and in another of horizontally and verti- 

 cally crossing threads, forming the reference-points by their inter- 

 section. In yet another tube an artificially fixed star is formed by 

 throwing gaslight through an exquisitely minute aperture, which 

 has its own image reproduced in the focus of a convex lens, so 

 contrived as to fidl only one-fifth of an inch from the curved surface 

 of the glass, a proceeding which practically reduces the image of 

 the simulated star to the very smallest dimension to which it is pos- 

 sible to compress it. The four reference-points, or collimators, are, 



