LYMAN. — TIIE SPECTRUM OF IIYDR0G1 EN. 133 
amount of washing necessary to secure it remains of course true. To this is now- 
added the trouble of pumping and filling the discharge tube with whatever gas may 
be under examination. The gas in the discharge tube, however, may be used over a 
considerable range of pressure, for the presence of the fluorite window prevents the 
discharge from spreading into the spectroscope. In practice the pressures varied in 
different experiments from 2 mm. to .5 mm. The pressure in the receiver was usually 
reduced to 0.1 mm., though if the hydrogen be pure so low a pressure is not at all 
necessary. The width of slit used varied from .09 mm. in the case of the crude plate 
published in a former article to about .025 mm. in the case of the fine plates from 
which measurements have been made. 
The time of exposure for the hydrogen spectrum varies between five and thirty 
minutes according to the width of slit and the sensitiveness of the plate. 
DRY PLATES. 
Little can be added to Schumann's description 3 of the manufacture of the spec 
dry plates necessary in this work. The first part of the research was carried on wi 
plates prepared from glass 1.5 to 2 mm. in thickness. When the form of the spect 
scope was 
special sheets of thin glass were required. In order that the emulsion may flow evenly 
the plates must be very flat. This necessity of flatness together with the mech 
improved and it became necessary to bend the plates to the arc of 
difficulties of grinding put a limit 
of the sheets. In practice plat 
I 
8.8 x 13 cm. and between .4 and .5 mm. thick are flowed and when dry are i 
small pieces 2.6 x 4.4 cm. 
One slight departure from the method of Schumann has been found ad 
each plate was separately supported on legs during the process of flowing. 
way if the emulsion run over the edge of one plate, only that plate is spoiled ; while 
if all the plates are on one levelling table a disaster to one may result in the over- 
flow of all. 
In development the author has used ortol with good results. Here, as has been 
remarked by Schumann, the strength of the developer must be regulated by the age 
of the plate. The addition of ice is a very necessary part of the process. The follow- 
•tions are suitable to plates six months to a year in age : Ortol A, 1 part, 
B 2 parts, Water 2 parts, ice about 1 part. 4 
» Ann. der Physik. Vol. 5, p. 349. 1901. 
4 A . Wa ter 1000 cc B. Water . . 2000 cc. 
Metabisulphate of Potash 7.5 gr. Pot. Carb. 120 grs. 
Ortol ...... 15 gr- Sod. Sulphite g» 
Hyp©. 1 to 20 20 cc. 
ing propo 
