E. ACKERMAN, G. K. STROTHER AND R. L. BERGER 29 



with a stopped flow method if the half times were greater than 5 msec or 

 with an accelerated flow method for half-time from 0.2 to 5 msec. In the 

 former case, the reaction curve was recorded on the Bruel-Kjaer appara- 

 tus. For accelerated flow determinations, the velocity of flow was dis- 

 played on the X axis of the oscilloscope and the optical density on the y 

 axis. A long persistence screen was used; the curves were either photo- 

 graphed with a Land camera, or drawn with wax pencil over the protective 

 grid, measured and cleaned oft". In order to vary the hydrogen peroxide 

 concentration over a factor of three hundred, it was necessary to use both 

 stopped flow and accelerated flow measurements, since ti/2 varies by a 

 similar factor. 



For most of the flow system experiments, the total optical density change 

 during the reaction was about 0.00015 with a 0.1 fiu catalase solution. 

 This is smaller by a factor of roughly twenty than the corresponding 

 change using cuvettes since the path length is shorter by this factor. The 

 tungsten lamp was operated from a stabilized power supply similar to 

 one described by Chance (6). The noise level under ideal operating con- 

 ditions corresponded to an optical density change of 0.00001 or less. Ideal 

 operation meant, among other things, doing experiments only from 1 to 

 5 A.M. Most of our experiments were made between 12 noon and 12 mid- 

 night ; since the noise level was appreciably higher at this time, it limited 

 the accuracy. 



EXPERIMENTAL RESULTS 



In the preliminary set of experiments the reactions were followed only 

 with the flow system at both 405 m/i, and 230 m/z. Armour powdered beef 

 liver catalase was used. The viscosity and dielectric constant of the me- 

 dium was varied with glycerol, sucrose and methylcellulose. All three 

 slowed the formation of p at 405 m^, increased the ratio of Pi/e, but left 

 the reaction at 230 ni/A relatively unchanged. The data at 405 niyu, led to 

 the conclusion that kj was independent of viscosity or dielectric constant, 

 and that k4 was also independent of dielectric constant but was diffusion 

 controlled. The data from 230 ni/x agreed with results that ki and k4 were 

 independent of the dielectric constant and that k4 was diffusion con- 

 trolled. However, these results indicated that ki increased as the diffusion 

 rate decreased. 



We felt it was necessary to repeat these experiments using a more puri- 

 fied catalase preparation. For all the results shown in the graphs and dis- 

 cussed below, Mann Laboratories crystallized beef liver catalase was 

 used. Here the results showed a consistent systematic difference of about 

 a factor or two in ki and k4 as computed from data at 405 as compared 

 with data at 230 m^. However, the percentage variation with temperature 



