DEGLI SPETTRv,SCOPISTI ITALIANI 39 



metre) long; then to this a scale was attached in the nianner to be described he- 

 reafter. From this compound photograph the upper half of the accompanying piate 

 was produced. Theu the origiual negative was inagnified about twice, aad from part 

 of tbat enlarged spectrum the lower half of the piate was made. The two spectra 

 are absolntely uuretoached, and are the work of the Sun alone. The ouly portion 

 where the haud of the draughtsmau is seen, is in the scale which is placed over 

 the upper or sraaller spectrum , and this is mostly copied from Angstrom's charti 

 Physicists may therefore rely entirely upon the correctness of the relative positions 

 and numbers of the lines and bauds which are shown. (Vedi Tav. XXXVI). 



The collodion negatives were next brought into contact with a film of gelatine 

 containing bichromate of potash, and a photo-lithographic piate was produced, from 

 which copies in printing ink could be made in a priuting press. This latter pro- 

 cess is known as the Alberttype , and for carryiug out the details of it I am un- 

 der great obligations to the owner of the patent, M.' Bierstadt, who has thus ena- 

 bled me to distribnte impressions, wich are not liable to fade like chloride of Sil- 

 ver photographs, and are as well defined as any paper copies of a glass negative 

 can he. 



The picture of the spectrum being therefore of unimpeachable correctness, it is 

 necessary to describe the method by which I attempted to give precision to the 

 scale which iudicates the wave-lengths of the ultra-violet rays. 



o 



I have taken as a basis the numbers given by Angstrom in bis celebrated Spectre 

 Normal du Soleil, for the wave-lengths of Cr, 64 and D^. If these numbers should 

 on future remeasuremeut be found to be slighlly erroneous, my wave-lengths will 

 require a proportionate correction, which can easily be effected. The method of nsing 

 these numbers was as follows. 



The line G of conrse can impress itself on collodion, and as it falls near one end 

 of my photograph, it furuishes a fixed point for the less refrangible portion of the 

 picture. But J^ and Dj do not He in that region of the spectrum which decompo- 

 ses the iodide and broniide of Silver in the common collodion and albumcn proces- 

 ses, and hence a particular description of the way in which these rays were uti- 

 lized is necessary. 



Some of the phenomena observed on passing a beam of light from a narrow slit 

 through a diffraction grating, and a suitable combination of lenses, may be thus de- 

 scribed. On a screen, placed on the side of the grating opposite to the slit, there 

 is seen a bright white image of the slit, and on citber side of this imago a series 

 of spectra, named in the order of their occurrence, spectrum of the P' order, spec- 

 trum of the 2"*^ order, etc. The spectrum of the P' order on each side is isolated, 

 but the red end of the 2^*^ order overlaps the violet end of the V*^ order. This over- 

 lapping gives great facilities for determining the wave-lengths of the ultra-violet 

 rays, for it tnrns out that if in the visiblc spectrum of the 2"^ order two steel points 

 are arranged, one to coincide with Dj, and one with è», on taking a photograph 



