42 



Mr. G. M. Whipple. On the Examination of [Feb. 15, 



II. " Description of an Apparatus employed at the Kew Obser- 

 vatory, Richmond, for the Examination of the Dark Glasses 

 and Mirrors of Sextants." By G. M. Whipple, B.Sc, 

 Superintendent. Communicated by Warren De La Rue, 

 Esq., Vice-Chairman of the Kew Committee. Received 

 February 6, 1883. 



In the "Proc. Roy. Soc," vol. 16, p, 2, Professor Balfour Stewart 

 described an apparatus designed and constructed by Mr. T. Cooke 

 for the determination of the errors of graduation of sextants. 

 This instrument has from that date been constantly in use at the 

 Kew Observatory, and since the introduction of certain unimportant 

 improvements, has been found to work very well. 



No provision was made, however, for its employment in the deter- 

 mination of the errors of the dark shades used to screen the observer's 

 eyes when the sextant is directed to the sun or moon, and it has been 

 found that errors may exist in the shape of want of parallelism in 

 these glasses, sufficiently large to seriously affect an observation, 

 accurate in other respects. 



It bas also been found that sextant makers are desirous of having 

 the shades examined before proceeding to fit them into their metal 

 mountings, and also to have the surfaces of the mirrors tested for 

 distortion before making the instruments up. With a view to the 

 accomplishment of these ends, for some time past the Kew Com- 

 mittee have undertaken to examine both dark glasses and mirrors, 

 and to mark them with a hall-mark, when they are found to answer 

 the requirements necessary for exactitude. 



For these purposes the apparatus now described has been devised by 

 the author, and brought into use at the Observatory. 



It is represented in the annexed cut. 



A telescope, A, of 3| inches aperture and 48 inches focal length, 

 a pair of collimators, B and C, of \\ inch aperture and 10 inches 

 focal length, and a heliostat, D, are firmly fixed to a stout plank, so 



