44 ROBERT RIKMENSPOEL 



For the total number, n, of cells (of all velocities) that pass the 

 aperture per second, 



n = c-v-Q (5) 



where c is the concentration of the sample and Q, is a proportional 

 constant. Since v is independently determined, equation (5) enables 

 one to measure the concentration. The constant Q was calibrated 

 by comparing the number of counts of a sample with the concentra- 

 tion measured from a film of the same slide : 



Q = 15.5 ± 1.2 /* 



The standard deviation in Q of 8% is caused by variation in sperm 

 size, focusing errors, etc. Since, in general, the statistical spread of n 

 is from 6 to 8%, the concentration can be measured in this way to 

 about 10 to 12% accuracy. 



Recently a signal analyzer was built, based upon the following 

 principle: the time, At, between two light flashes of the head can be 

 written, 



At = p-l/v 



where p is the constant in equation (1). Whereas p has a scattering 

 (standard deviation) of about 25% for different cells the average 

 value of p has a much smaller spread of about 6% from one sample 

 to another (see Rikmenspoel, 1957b, p. 67). Thus, useful data can be 

 expected by equating: 



V = p-l/At 



and taking for p the average value, p = 1 1.9 microns/sec. 



This principle has been technically realized in the following way. 

 By means of a discriminator, the spikes caused by the sperm head are 

 transformed into short pulses. The first of these pulses sets into ac- 

 tion a string of univibrators, the delay time of which is adjusted so 

 that each represents a velocity class of 20 microns/sec interval. The 

 delay times should be thought of as in series. A scale of 2 selects the 

 second flash. Coincidence circuits determine with which univibrator 

 this second flash is coincident; mechanical counters register the num- 

 ber of observed cells in each class. In this way a velocity distribution 

 is immediately obtained. The total number of cells passing the aper- 



