420 Prof. D. B. Brace on a New Spectrophotometer 



tive discharge, is the fact that k is probably less for blunt 

 than for sharp points with a given current. 



The smaller the current, however, the smaller this dis- 

 crepancy will be ; and a small current is also likely to give 

 rise to a smaller back-discharge error. On the other hand, 

 the error which is due to the luminous projection from the 

 negative point may be reduced by using a sharper point. 

 Two points of respective diameters 0'013 cm. and (approx.) 

 O001 cm. were therefore selected, and sixty observations of 

 the 7r-pressure made as before with each for an average 

 current of 990 e.S. units. Taking k=\-5 the number of mole- 

 cules in the diameter of a positive cluster deduced from these 

 data was 16 ; that for the negative cluster 12. These numbers 

 are consistent with the higher speed of the negative ions, 

 besides being likely to be more reliable than those given 

 above for the reasons just stated. Whether the use of still 

 smaller points and currents, if it were practicable, would alter 

 them much it is impossible to say. At least they represent 

 the order of magnitude of the quantities concerned, and they 

 thus afford considerable support to the theory of molecular 

 clusters. 



In conclusion I wish to express my thanks to Mr. E. H. 

 Dixon for valuable aid during the later experiments on the 

 estimation of 7r; and to my assistant, Mr. Herbert Strachan, 

 who helped me with the greater part of the work, and to 

 whose skill and patience much of the success of the experi- 

 ments is due. 



XLV. Description of a New Spectrophotometer and an Optical 

 Method of Calibration. By Prof. D. B. Brace, Ph.D.* 



TO Govi, Vierordt, Orova, Cornu, Glazebrook, Wild, 

 Konig, and Lummer and Brodhun, and others are 

 due different types of spectrophotometric instruments. The 

 arrangement adopted by Govi, Vierordt, Crova, and others of 

 bringing the two spectra to be compared into contiguous 

 positions by means of double slits and total-reflecting prisms 

 fails in eliminating the dark line separating the two, and thus 

 reduces the sensibility of the eye in determining a match, as 

 is well known to be the case unless the bounding lines of the 

 two fields are perfectly sharp, and vanish completely when 

 set for a match. The same defect exists in the polarizing 

 arrangements of Glan, Crova, Glazebrook, Konig, and others, 

 where the two spectra are brought into juxtaposition by 



* Communicated by tho Author. 



