ON SPECTR OS COP Y. 13 



the metallic spectra, showing which lines of these latter are coincident 

 with dark lines in the solar spectrum and which are not, and thus 

 confirming the presence of elements suspected to exist in the sun's 

 reversing layer by previous observers, and proving the presence of 

 numerous others. Moreover, there are considerations which point to the 

 existence of solar elements which are quiteforeign to, or have not yet been 

 discovered in this world. Mr. Lockyer exhibited the maps which have 

 been constructed for this purpose, and, up to the present time new maps 

 have been constructed for the region of the solar spectrum comprised 

 between wave-lengths 39.000 and 43.000. The metallic spectra 

 for the whole of this region are not yet completed, as the amount of 

 labour involved in their production is very considerable. However, 

 the spectra of about one-half of the metals for the first two sections, 

 viz., wave-lengths 39.00-40.00 and 40.00-41.00, are nearly completed, 

 and these are, in all probability, absolutely pure. The method of 

 purification was then stated. Here the lecturer observed that he 

 was glad to have that opportunity of expressing his obligations 

 to Corporal Murray of the Royal Engineers, who had most efficiently 

 prepared the enlarged maps, which were finally photographed 

 down to the size desired. The enlarged maps are drawn on twelve 



O 



times the scale of Angstrom, and the final photograph will be re- 

 duced to four times, this being none too large to ensure the 

 necessary amount of detail. To show how vast an improvement the 

 photographic map exhibits over that of Angstrom's, Mr. Lockyer stated 

 that whereas in Angstrom's map there were only three lines between 

 the H i and H 2 lines (of the solar spectrum), the present maps showed 

 ninety-nine. The lecturer proceeded to give some details of the 

 manner in which photography had been utilized in the research. The 

 instrument used consists of a spectroscope constructed on the model of 

 Bunsen and KirchofF s, with a train of four flint-glass prisms, 3 of 45 

 and i of 60. The observing telescope is replaced by a camera pro- 

 vided with a simple quartz lens of about 6 feet focus. By these means 

 the image of a tolerably large portion of the spectrum is thrown upon 

 the photographic plate at the end of the camera, and a good focus for 

 a part of the spectrum, at any rate, ensured. The novel feature of 

 Shis spectroscope consists in the adaptation of a sliding shutter to the 

 slit. It is well known that the width of a spectrum depends upon 



