190 COLLOIDS IN BIOLOGY AND MEDICINE 



complement is inactivated, since it may be reactivated by addition 

 of "end piece" and of "middle piece." Serum becomes turbid on 

 shaking, and it is the author's opinion that this is evidently due to a 

 coagulation of the albumin by shaking. 



In many investigations, especially in "immunity studies," it is cus- 

 tomary to shake the test tubes, and I believe that some of the dis- 

 agreements in the results of experiments by different investigators 

 are due to a disregard of such surface phenomena. [R. OTTENBERG 

 does not consider this factor in his exhaustive study. Arch, of Int. 

 Medicine, vol. xix, pp. 457-492. Tr.] 



The diffusion coefficient of several enzymes was measured l by R. 0. 

 HERZOG and H. KASARNOWSKL* They are for 



Pepsin 0.062 (at 12 C.) 



Pepsin 0.066 (at 16 C.) 



Rennin . 062 (at 16 C.) 



Invertin 0.032 (at 16 C.) 



Emulsin 0.033 (at 15.3 C.) 



From these figures the following molecular weights were cal- 

 culated: 



Pepsin 13,000 



Invertin 54,000 



Emulsin 45,000 V 



In studies on the filtration, ultrafiltration and diffusion of enzymes 

 through membranes it must be determined beforehand, whether the 

 filter adsorbs too strongly. Thus, e.g., a Chamberland filter per- 

 mits no pepsin, trypsin, lipase or zymase to pass through, though 

 the pores are of ample size. By choosing suitable membranes, 

 these methods of separation have given valuable results. It has 

 been possible by diffusion and ultrafiltration to separate a num- 

 ber of enzymes, which were formerly regarded as individual, into 

 two constituents having different properties. Thus, according to 

 S. FRAENKEL and M. HAMBURG,* diastase prepared from malt may 

 be divided into two enzymes. The one which diffuses changes 

 starch into sugar, whereas the other merely fluidifies the starch. 

 A. VON LEBEDEW* ultrafiltered expressed yeast juice and thus 

 succeeded in demonstrating, that in fermenting sugar the disap- 

 pearance of the sugar and the formation of carbonic acid are two 

 distinct processes. 



In the course of such experiments it has been shown frequently, 

 that the components are inactive individually and only exert their 

 enzyme action in combination. The first observation of this kind 

 was that of R. MAGNUS,* who dialyzed liver extract. The extract 

 which originally split up fat thus became inactive; when MAGNUS 

 1 The figures are the mean of several determinations. 



