and Laboratory Methods. 1948 



membrane of parchment paper is cut to the outer diameter of the flange, and the 

 position of the screw holes marked with a pencil ; openings opposite to the holes 

 are punched out, and the membrane is wetted and laid evenly over the rubber 

 ring ; then the screws are inserted and screwed home, and the receiver filled 

 with the solution. Connection is next made with the manometer by means of a 

 piece of barometer tubing of thick wall and narrow bore. Parchment paper was 

 used as a membrane. The authors found this to be absolutely impermeable to 

 colloids, and in other ways very satisfactory. 



The colloid most used in the experiment was sheep's serum. The osmotic 

 pressure of a sample of this was first obtained, and then the serum albumen was 

 converted into alkali albumen by adding a sufficient quantity of ten per cent, 

 caustic soda solution to make a one per cent, solution, and boiling. The same 

 percentage amount of caustic soda solution was then added to the outer fluid, 

 and readings were again taken. A much higher pressure was invariably obtained 

 in the second case, indicating that the size of the solution aggregate in the alkali 

 albumen was smaller than in the serum. In a typical experiment a reading of 

 22 mm. of mercury was obtained with serum, while with alkali albumen the 

 osmotic pressure was equal to 110 mm. of mercury. Thus the conclusion that a 

 definite osmotic pressure is exerted by colloids in solution is established. The 

 authors are inclined to think that the differences in properties of coagulable pro- 

 teids may be due to different physical arrangements of chemical molecules to form 

 different aggregates, and that the complexity of the proteid molecule may be 

 much more a physical than a chemical phenomenon. 



For an account of the numerous experiments, as well as for the valuable and 

 interesting discussion of the important bearing of the results of this work on 

 general physiological problems, recourse must be had to the original paper. 



R. p. 



NORMAL AND PATHOLOGICAL HISTOLOGY. 



JOSEPH H. PRATT, Harvard University Medical School. 



Books for Review and Separates of Papers on these Subjects should be Sent to Joseph H. Pratt, 

 Harvard University Medical School, Boston, Mass. 



Wright and Joslin. Degeneration of the Is- Schafer, Diamere, Ssobolew, and Lag- 

 lands of Laneerhans of the Pancreas in , , , .1 „. .1^ ;^i„„j„ 



Diabetes. The J. of Med. Research, 2- "^sse have suggested that the islands 



360-365, 1901. of Langerhans might play an important 



part in carbohydrate metabolism. Opie has given this idea strong support. 

 Among fourteen cases of chronic interstitial pancreatitis the islands of Tanger- 

 hans were found to be the seat of hyaline degeneration in four. In three of 

 these four cases diabetes mellitus had been present, but had been absent in the 

 others. In the fourth case the islands were only slightly affected. Moreover, 

 in two cases of diabetes mellitus Opie found there was no increase in the inter- 

 stitial connective tissue of the gland, but the epithetial cells of the islands of 

 Langerhans were more or less extensively transformed into hyaline material. 



