STUDIES ON CHROMOSOMES. 25! 



with the necessary degree of accuracy. It is not always easy to 

 determine the exact base or the exact apex of the head. The 

 most vexatious fact was that the long heads of nearly all of the 

 spermatozoa were slightly curved and it required much searching 

 to find ones sufficiently straight to measure accurately. Other 

 likely sources of error were guarded against as carefully as 

 possible. As nearly as could be done exactly the same fixation 

 and degree of staining was secured in each slide. Care was 

 taken to see that the film of seminal fluid was spread on in 

 approximately the same degree of thinness and that only such 

 parts of the preparation were studied as appeared to be wholly 

 free from stretched or distorted spermatozoa. To eliminate the 

 personal equation, as already related, another person was ob- 

 tained to do two series of counts. Still other sources of error 

 such as might arise through foreshortening of the object to be 

 measured, or fatigue on the part of the observer, were guarded 

 against. 



A Leitz Stufen-Mikrometer in a No. 2 eye-piece was employed 

 and the readings were based on measurements down to the mid- 

 field between the individual lines. The lines, as is usual in 

 micrometers, were laid off in tens. Inasmuch as relative sizes 

 were all that was desired the figures accompanying the various 

 curves represent merely divisions of the ocular micrometer. It 

 is evident (Text-fig. 2) that three of the curves show no trace of 

 bi-modality. A slight trace appears in b, but so slight as to be 

 negligible since the curve is based on measurements of only one 

 hundred spermatozoa and the actual differences between the 

 10.5 group and the II group is only three individuals, there being 

 20 spermatozoa which measured 10.5 and 23 which measured 

 II. Moreover B is a unimodal curve based on measurements of 

 515 spermatozoa from the same individual as that from which 

 those measured in b were taken. With the differential X-like 



B. frequency distribution of head-lengths of 515 spermatozoa from a second Ply- 

 mouth Rock fowl. 



Scale value 7-5 S 8.5 9 9.5 10 10.5 n 11.5 12 12.5 13 13.5 



Frequency i 23 7 40 188 169 90 10 2 o 2 i 



b, frequency distribution of head -lengths of 100 spermatozoa from the same fowl 

 as in B, in a different preparation and by different observer. 



Scale value 9 9.5 10 10.5 n 11.5 



Frequency i 1 1 38 20 23 7 



