spermatozoa was not added up until all the slides from the five
replicates of an experiment had been counted. Two observers counted
all the slides.

The significance of the results was assessed by analysis of
variance of the percentages and after transforming them to angles,
the transformation did not alter the results of the analysis.

The examination of rabbit semen was less comprehensive. The
same range of temperatures was used but undiluted semen was only held
at the experimental temperature for 5 minutes before being stained at
the same temperature and diluted samples were held for 10 minutes and
then stained at 30°C.

RESULTS. Ram. The same proportion of eosinophilic spermatozoa was
counted in samples kept at 20°C and 30°C but at 10°C, 7.5°C and 0°C
an increasing number of spermatozoa were eosinophilic. The propor-
tion of eosinophilic spermatozoa was higher after dilution (fig. 1)
but the time of exposure and the temperature of staining had no
effect.

Bull. The results for bull semen are essentially the same as
for ram except that diluting bull semen did not produce a higher
proportion of eosinophilic spermatozoa (fig. 1) whereas samples
stained at 30°C had a higher proportion of eosinophilic spermatozoa
than those stained at lower temperatures.

An examination of results for different replicates (i.e. differ-
ent ejaculates) showed that the difference between staining at 30°¢
and at the lower temperature could be accounted for largely by
extraordinary variations in the response of samples at 7..5%C4

Rabbit. The proportion of eosinophilic spermatozoa did not
vary significantly between treatments.

DISCUSSION. In some circumstances it is easy to distinguish between
eosinophilic and non-eosinophilic spermatozoa in the nigrosin-eosin
stain but under other conditions (5) the distinction is not so clear
cut. When these conditions prevail it is possible for decisions by
the observer to be unconsciously biassed, coding the slides reduces
this possibility. Decisions may also be influenced by observers un-
consciously attempting to reduce the variability within a sample (6),

(5) H. M. DOTT, 1956. III Int. Cong. Anim. Rep. & A.I. Cambridge.
(6) R. C. CAMPBELL, H. M. DOTT & T. D. GLOVER, 1957. J. Agric. Sci.
48: 1.

1236