128 LYMAN. — THE SPECTRUM OF HYDROGEN. 
The apparatus is so constructed that the curve to which the plate is bent passes 
li the slits. Light has access to the plate through a slot T cut in the disc, 
o"- u "~ " — ■ -~-o 
which slot also serves as a diaphragm for the spectrum. A sleeve, U, shields the plate 
from scattered light ; and to reduce the reflection from the walls of the tube a set of 
circular diaphragms are provided. The whole system, draw tube and large tube, are 
blackened inside by the usual process. In the early work it was proposed to enclose the 
spectroscope as above described in a large glass tube, but owing to the difficulty of clos- 
ing such a receiver air-tight, and owing to the great liability of tubes of this size to 
break, the plan was abandoned. The receiver at present in use consists of a drawn 
br; b tube 11.3 cm. in diameter, 110 cm. long, and 1.8 mm. thick. It is provided with 
two flanges, one at each end, cut from sheet brass and soldered to the tube. The flange 
at the end destined to be nearest the grating is closed by a circular brass plate, ground 
true, some 17 cm. in diameter. Plates of two kinds have been used to close the other 
end of the receiver. In the simpler form shown in Plate VII, fig. 2, a circular brass 
disc was only pierced by the two holes destined to admit light to the slits of the spec- 
troscope. In the more complex form, fig. 4, a hand hole is also provided through 
which the plate-holder may be introduced. This hole is 6.2 cm. in diameter and 
is closed air-tight by a conical plug. In order to give this plug a sufficient bearing, 
a sleeve some 4.5 cm. high is attached to the face plate. An inlet tube inserted about 
midway down the length of the receiver serves to exhaust the air; a wooden frame 
holds the apparatus horizontal. To facilitate the handling and development of the dry 
plates the end of the receiver is inserted in a small dark room. Plate VII, fig. 1, shows 
the appearance of this arrangement. Into the receiver thus described the spectroscope 
is slipped, small hard rubber-legs hold it in a central position. Plate VII, fig. 3, shows 
the end of the apparatus with the face plate removed. 
^ The concave grating with which the work has been done was ruled in 1903 on the 
improved engine at Johns Hopkins University; The material is the usual speculum 
metal, the radius 97 cm. ; there are 15,028 lines to the inch. The diamond point was 
selected with the object of throwing as much of the light as possible into one spectrum, 
lo the great success which attended this effort the results of the work are due, for the 
instrument possesses one first spectrum of extreme brilliancy. 
s the experiment is carried on in an atmosphere of hydrogen the preparation of 
ie gas 18 nt factor> Zbc and hydrochlor . c add Qf ^^ _ 
EllTitr* "^ The *» is P-ed over potassium hydrate and 
water. Leiore the gas is admitted to the spectroscope it is dried 
cnlonde and phosphorous pentoxide. The drying tubes are protected at 
greatest com- 
lected over distilled 
