179 



explained lo him verbally by me just before the experiment. We 

 both worked in the same room but we had absolutely no influence 

 on one another's results. My colleague got the following values for 

 x in 11 determinations: 7.2, 6.9, 7.1, 6.8, 7.05, 6.7, 6.7, 6.65, 6.7, 

 6.85, 6.65; average, 6.845, mean error, 0.185 = 2.7 o/ . I obtained 

 the following for x in 12 determinations: 7.35, 6.65, 7.1, 7.0, 6.9, 6.9, 

 6.9, 6.75, 6.85, 6.65, 6.95, 6.9; average, 6.91, mean error^ 0.19 = 2.7o/ . 

 This degree of accuracy must be considered particularly satisfac- 

 tory. Errors in pipetting, measuring the diameter of the tubes elc. 

 are even included in the mean error; the mean error of the opti- 

 cal determination alone can hardly be put at more than 2o/o. 



A third investigator who was not used to estimating the con- 

 centration of suspensions, obtained divergent results, namely in the 

 first series: 8.65, 8.65, 8.55, 8.45, 7.95, 8.4, 8.6, 7.6, 7.2, 7.5, 7.4, 6.4; 

 and in the following series (after having been assured that these 

 figures were on an average „too high"): 6.4, 6.6, 6.5, 6.2, 6.8, 6.6, 

 6.3, 6.3, 6.2, (6.43 + 0.21 (3.3o/ )). These divergent results, - - psy- 

 chologically easily intelligible, — of an unpractised worker, can 

 however hardly depreciate the value of the method. 



The concentration determined in a test tube of 15 mm's diameter 

 by the method described, is used in the direct agglutinations, hut is not 

 defined as the unit of concentration. As concentration 1 is defined 

 that concentration through which letters can just be distinguished 

 when the tube (which should not have too thick walls) is 1 cm. in 

 internal diameter. In the agglutination experiments therefore suspen- 

 sions of a concentration about 2 / 3 were used and consequently the 

 concentration in the mixture of culture and serum dilution was 

 about 0.4. 



In a couple of experiments with a suspension of a weighed 

 amount of a surface culture (of semi-fluid consistence) it was found 

 that the concentration 1 corresponds to 5—6 mg. per c.c. 



If it is desired to prepare a suspension of the concentration n 

 (n > 1), a suspension is first made which is obviously more con- 

 centrated. It is determined how many times (m) a sample of it 

 must be diluted to get a concentration 1. The remainder there- 

 fore is to be diluted — times to obtain a suspension of the con- 

 centration n. 



If the concentration desired is n < 1, a suspension of the con- 

 centration 1 may first be prepared, which is then diluted to — times 



the original volume. 



Concentration determinations can be made not only with daylight 

 but also with lamplight. In the latter case it is advisable however, 

 always to use the same source of light, to place the tube always 

 in the same relative position to it, and compare the determinations 

 made by lamplight with those made by daylight. 



The unit which is determined by clear diffused daylight a 

 couple of metres from an ordinary window, should serve as the 

 standard. :"•■!, 



