TIIK AllSOUn'lUN AMI KM1S.S1UN UK AIK. 



Fig. 1. 



hi,:. I- 



/■'.<■- 



^^ 



o. 



la © 



fine adjustment of the slit to iiaiallelisin witli tLe refiaotiiig edge ot" the prism. 



The three micrometer screws ai'e rendeied air-tight by means of the three steel 



coues, to", s", 10". The slit box d" is 

 connected with the source of Hgbt 

 tlirough a cover shown in Fig. 5, e". 

 This cover, e", carries by a flat 

 ground greased joint the Geissler 

 tube, which is centered by the steel 

 pin ("', and after centering is secured 

 iMGs. 1-3. rn.noGRAiMiicPLATE-Hoi.DER. NATURAL sizE. agaiiist lateral shifting by a slide, 



h". It is needless to say that the 



steel pin is withdrawn when the ceuteiing of the tube has been effected. The 



mode of connection of the tube with the vacuum-spectrograph does not permit the 



two to be filled independ- 

 ently of one another with 



such gas as may be under 



investigation at the time. 



It will not do to fill both 



unless the gas is suffi- 

 ciently transparent. This, 



howevei', is not the case 



with the majority of gases ; 



so that a window has 



usually to be interposed 



between the Geissler tube 



and the cover, e", so as to 



diminish the thickness of 



the stratum of gas to be 



traversed by the rays. The 



same arrangement is use- 

 ful in pliotographing by 

 sjiarks and arc-light. In 

 case the observation is not 

 to extend farther to shorter 

 wave-lengths than 155 f^^t, 

 n sufficiently thin (juartz 

 window will answer the 

 jiuritose. Foi' moi-e re- 

 frangible rays it is not 

 transpai'fint enough, and 

 white fluor-spar is tlie only ri(. 



suitable matei'ial. 



Fig. 6 shows a section of tlie part in which the prism (which appears in B'ig. 

 b as k) is set up. A thick-walled cone, b, is rigidly connected with the collimator. 



Si 1 r WITH THREE MICROMETERS. NaTIRAL SlZE. 



