414 



Scientific Proceedings (131) 



present. Heating the first filtrate (containing much protein) 

 at P H of 4.0 does not always produce a coagulum and if not, may 

 increase the potency. 



202 (2162) 



Precipitation reactions of insulin. 

 By C. P. KIMBALL, R. S. ALLEN and H. A. PIPER. 



[From the Physiological Laboratory, University of Rochester, 

 Rochester, N. Y.] 



We here report the results obtained by the addition of various 

 reagents to the aqueous solution of insulin as it is prepared for 

 injection in human cases. Potency tests were made on normal 

 rabbits, the dose given being equivalent to 20 gms. of pancreas 

 except in a few of the earlier experiments where the equivalent 

 of 40 gms. was taken. Only those instances where some positive 

 test was obtained with the same material treated at the same time 

 are reported. 



Up to date 29 reagents embracing a wide range in chemical 

 nature, have been tried of which two (formaldehyde and ethyl 

 acetate) gave no precipitate of any kind. Next comes a list of 

 6 which gave a definitely negative result, i. c, contemporary ex- 

 periments with the same material gave some positive tests with 

 other reagents. It should here be explained that on the addition 

 of the reagents the precipitate is thrown down by centrifuge- 

 taken up in sterile water and injected, the supernatant liquid 

 is dried down, with or without dialysis, depending on the nature 

 of the precipitant, and similarly dissolved and injected. The 

 following 6 gave definite negatives : ether, petroleum ether, 

 toluene, xylene, chloroform and cadmium chloride. Five more 

 are probably destructive to the principle, phosphotungstic, phos- 

 phomolybdic and tannic acids, MgS0 4 and NaS0 4 although the 

 evidence is not yet conclusive. Mechanical difficulties were en- 

 countered with phenyl hydrazine, pyrogallic and picric acids, 

 which have eliminated them from our list. Too little has been 

 done with U0 2 Ac 2 , Zn S0 4 and NaCl to justify any inferences. 



Turning now to the positive results we find that absolute ethyl 

 alcohol is unique in being the only reagent encountered so far 



