296 SECTIONAL TRANSACTIONS.— A. 



{h) Dr.'H. AV. B. Skinner.— .4 Two-Metre Vacuum Spectrograph. 



The objects kept in view in the design of this instrument are (1) ease of adjustment 

 and (2) range of utility. The first point is attained by taking all the adjustments out 

 through the wall of the main tube of the spectrograph by means of small ground joints 

 and arranging a lighting system within for observing the various scales ; also by arrang- 

 ing that the plate may be removed through the side of the main tube through a large 

 ground joint without the necessity of removing the heavy end-plate. The second 

 point is satisfied mainly by using as large a photographic plate as convenient, and 

 allowing range of adjustment so that the spectrograph may be used up to the visible 

 region. Actually the plate is 27-5 cm. in length and the wave-length regions 0-2000 

 A.U. 2000-4000 A.U., &c. , may be photographed on separate plates. This, of course, 

 implies a large spectrograph tube (35 cm. dia.) and a correspondingly fast pump for 

 evacuation. 



The spectrograph is carried entirely on a steel girder which lies within the copper 

 spectrograph tube. At the one end the girder carries the grating mount which is 

 carried on three levelling screws with tangent screw adjustment and length adjustment. 

 The plate-holder and slit are carried on the other end of the girder. The whole may be 

 removed as a unit from the copper tube. The plate-holder may be raised or lowered 

 from outside the main tube for the purpose of taking a number of exposures on a single 

 plate without re-evacuation. The plate is bent into a circle of two metres diameter 

 (this is possible with plates of special thin glass). It can be rotated by means of a screw 

 and bearing about an axis which coincides with the slit. Thus difierent spectrum 

 regions may be focussed with one adjustment of the plate (two adjustments of the 

 grating are also necessary). The slit is carried on a tube which projects through the 

 end-plate into the main tube. It is thus buried, but since it must lie on the Rowland 

 circle, this is unavoidable unless we adopt the rather complicated device of fixing a 

 subsidiary slit-tube inclined at a small angle to the main tube. It was thought better 

 to retain the slit as part of the girder unit in spite of the slight inconvenience due to its 

 burial. It can be adjusted from outside the spectrograph and ean be removed easily 

 for cleaning by means of a slip-tube. 



The end-plates are castings held on to the main tube b}'' bolts and made vacuum 

 tight by rubber and grease. A large Siemens-Schuckert pump with a speed of about 

 2 litres/sec. is used for evacuation. It will pump the 200 litres volume from atmospheric 

 pressure to 1 mm. in about 10 minutes. 



The detail design can be seen from the drawings. A Blythwood metal grating of 

 14,000 lines per inch and of size 8x5 cm. is at present in use. 



(i) Mr. G. Herzberg. — Investigations of the Band Spectra of P.^, C.j and CH 

 in the Ultra-violet and in the Vacuum Region. 



Though there is quite a good deal known about the structure of the nitrogen 

 molecule, so far hardly anything is known about the Phosphorus molecule (Pa) which 

 ought to be analogous to Na. Therefore the band spectra of Phosphorus are studied 

 here in greater detail. An extended band system of Po (below X 3500) has been found, 

 part of which was measured 20 j^ears ago b}' Geviter. There are some peculiarities in 

 the vibrational states of the two electronic levels involved which make an analysis 

 rather difficult. It is hoped to get plates with the two-metre vacuum grating in order 

 to resolve the fine structure which seems to be rather simple. In the long wave-length 

 region the bands extend only to a certain value of the rotational quantum number 

 (predissociation). Another band system, the origin of which is not yet quite certain, 

 has been found. The conditions of excitation point to CP as carrier. The fine 

 structure of the bands of P2 would also be interesting in relation to nuclear spin and j 

 statistics. 



Until now only very few band systems of Co and CH have been found. In order j 

 to test the validity of recent theories of molecular structure it would be highly interest- 

 ing to get knowledge about the higher electronic levels of these molecules. Therefore, I 

 their spectra are being investigated in the vacuum region. 



(j) Mr. G. Herzberg and Mrs. L. Herzberg. — An attempt to obtain^ 



Bands of Be.^ and B.^. 



According to the quantum mechanics of homopolar binding the formation of 

 molecules of Be and B should be possible though they have not yet been observed. 



