236 PROCEEDINGS OF THE AMERICAN ACADKMY 



inside the focus, witliout &lit or lenses; and witli this, after one or two 

 unsuccessful trials, he first obtained a satisfactory spectrum of Vega 

 showins four dark lines. 



Afterwards direct vision prisms used in the same way were tried, 

 and spectroscopes made up of such prisms, some with a slit, some 

 without, and some with a cylindrical lens to give jiecessary width to 

 the spectrum. The arrangement finally settled upon, however, and 

 with which all the plates measured by Professor Pickering were msule, 

 was the following. A star-spectroscope by Browning, with two CA)^ 

 prisms of dense {but white) Hint glass, was used, of the form designed 

 by Dr. Pluggins for stellar observations. The telescope and collimator 

 each had a focal length of 6 inches, with an aperture of | of an inch. 

 The parts were very carefully braced together to prevent any slip or 

 movement. The slit was covered with a diaphragm having a hole at 

 the centre, and painted with phosphorescent paint to make the aper- 

 ture visible in the dark : there was also a movable '• linger," by which 

 any part of the slit could be exposed at pleasure, so as to obtain spec- 

 tra of different objects on the same plate side by side for reference. 



At the eye-end of the si)ectroscope-telescope the eyepiece and 

 micrometer were removed, and a block of hard wood was fitted on iu 

 such a way as to carry the little photographic plate. This w^as a small 

 bit, about an inch s(piare, cut from a plate of commercial size. A 

 small positive eyepiece was mounted on the block, so that the operator 

 could at pleasure examine the yellow and red portion of the spectrum 

 which projected beyond the sensitive plate into the field of view, and 

 in this way assure himself that the clockwork was driving properly, 

 and that all the adjustments remained correct. 



The whole apparatus weighed less than five pound-, and screwed on 

 to the eye end of whatever telescope it was used with. 



On the most careful examination, it is very diflicnlt to see how any 

 percei)tible alteration in the relative position of the different j>arts 

 coulil ever have occurred. Still, the bracing employed was not abso- 

 lutely symmetrical, and there may have been a little " twist " when 

 the instrument was transferred from an object near the zenith on one 

 side of the meridian to one near the horizon on the other. 



For the most part the development of the plates was by ferrous 

 oxalate, though the alkaline development and pyrogallic acid were 

 both used on some occasions. 



The pictures obtained with this arrangement were about one six- 

 teenth of an inch in width and about half an inch long, extending from 

 a point between the Fraunhofer lines F and G lo a point near M. 



