928 BIOLOGICAL EFFECTS OF RADIATION 



Another method, employed with Uquid yeast, involving fewer steps 

 than the mycetocrit, is the nephelometric method. The yeast suspension 

 is diluted 1 to 10 and the difference in scattering and absorption of the 

 light is determined either by an automatic method (Frank, 75) or by means 

 of a Duboscq colorimeter (Klenitzky, 160). In the last method a calibra- 

 tion curve must be made for each set of experiments, that is, the setting 

 of the colorimeter must be tested with the different dilutions of a control 

 experiment and from these figures a calibration curve drawn. From this 

 curve the percentage difference can be determined easily. 



Since the yeast method is used in more than 90 per cent of the work 

 of Gurwitsch's school, it may be well to discuss here the merits and dis- 

 advantages of the different types of yeast-increase determinations. How- 

 ever, it is claimed that yeast may be used as sender or detector only in 

 the presence of visible light (213). A criticism of this procedure is made 

 later, in view of the factor of stray light. 



The yeast-bud-counting method from agar surfaces seems to be the 

 least "objective." Taking samples from the Petri dish before exposure 

 and counting buds and mother cells by the thousands are tedious tasks 

 and require great patience, since there are many possibilities for errors. 

 The liquid-yeast method seems to the reviewer the more objective. The 

 use of the blood-counting chamber is well established; but if it is not 

 standardized it may be a large source of error. For the reason just 

 stated, the mycetocrit technique would appear to add to the objectivity 

 of the liquid method; since the mycetocrit gives quick results and thus 

 permits the handling of a large number of experiments. The nephelo- 

 metric method, in the reviewer's experience, is not sufficiently sensitive 

 for the accurate determination of an increase of yeast. At least, it seems 

 that the use of the usual type of colorimeter is not so exact as the work 

 demands. 



In all this work it is presupposed that the yeast growth in two separate 

 test tubes is so uniform that the tubes could be used interchangeably 

 when not exposed. The presence of a mitogenetic effect is determined 

 by either an increase or decrease over the control. Many investigators 

 doubt that conditions can be maintained so perfectly uniform that 

 the growth of yeast will not be somewhat influenced thereby. Since 

 increase and decrease are accepted as effects, it would seem necessary 

 that all experiments should be run in duplicate, at least, and that all 

 controls check each other. That this is being done is not indicated 

 in the publications. It is reported by Gurwitsch (108) that each experi- 

 ment must be repeated at least eight times before it is accepted as 

 definite. It is, however, not unusual for a very important conclu- 

 sion to be made from two experiments performed at different times 

 (Klenitzky, 158). 



