328 



Dr. S. Judd Lewis. 



to determining the various characteristics of the absorption bands and to 

 finding how these may be accounted for. 



The subject may conveniently be treated in three divisions : — 



(1) Methods and Equipment. 



(2) Absorption Spectra of Blood Sera. 



(3) Review. 



Methods and Equipment. 



Most of the older methods of studying ultra-violet absorption spectra 

 depend on taking a series of absorption spectra, obtained on passing light 

 through layers of the substance varying in thickness or in concentration. 

 The modern method is to employ a sector spectrophotometer in conjunction 

 with a quartz spectrograph, and produce only one absorption spectrum 

 resulting on the passage of light through a suitable layer of the substance, 

 and to compare it with each member of a series of normal spectra differing 

 from one another by known amounts in intensity only. Thus, a normal 

 spectrum of one-third of the original intensity will match the absorption 

 spectrum at those wave-lengths where one-third of the light is transmitted, 

 that is, where two-thirds of the light are absorbed ; and similarly for any 

 other proportion. 



The apparatus employed were : (a) A large size (" size C ") quartz 

 spectrograph made by Hilger, 1914 model, (b) A sector spectrophotometer. 

 Two of these instruments have been employed in the course of the work. 

 The one referred to in the opening of the paper is that introduced by 

 Messrs. Adam Hilger, Limited. It is described by Lankshear in a paper 

 in the 1 Memoirs and Proceedings of the Manchester Literary and 

 Philosophical Society' (vol. 58, No. 15, Part 3, pp. 1-12, 1914). The second 

 instrument is a still later model made by Messrs. Bellingham and Stanley, 

 Limited. It differs from the Hilger in its construction providing for the 

 incident pencils of light being divided and brought together again by 

 reflecting prisms instead of by refracting units, and in the " sector " being 

 so constructed as to present an aperture which has the shape of a 

 goemetrical sector, and so to give a constant exposure instead of an inter- 

 mittent one. This avoids the necessity for correcting each batch of plates 

 for the error due to intermittency. (c) A powerful spark lamp provided ' 

 with nickel steel electrodes. 



The experimental procedure adopted for the serum work is as follows : — 

 A photograph of the two juxtaposed spectra is taken with both sectors 

 of the photometer fully open, and with nothing in either path save an 



