520 
ZOOLOGY: W. J. CROZIER 
the individuals forming a successful mating pair. Reciprocal fertili- 
zation may be presumed to occur in most of these cases, and is certainly 
carried out in some instances. 
The observation of mating pairs of Chromodoris zebra repeatedly 
suggested that under natural conditions this species forms copulating 
pairs of which the individual components closely correspond with 
each other in general size. Sexually mature specimens of this species 
range in length from 4 to 18 cm. Differences in the relative sizes of 
the various individuals are readily detected by the eye, while other 
dimensions of the animals may be subjected to measurement in the liv- 
ing condition. 
One hundred and forty-eight pairs of copulating C. zebra were ob- 
tained in the field. The total length (anterior edge of the mantle to 
posterior termination of the foot) was determined for each specimen, 
according to a method^ giving results sufficiently reproduceable for the 
purposes of statistical treatment. As a check upon this measurement 
certain other dimensions were ascertained, including the weight and the 
volume. Each of the methods of evaluating size yields the same qual- 
itative result. Therefore the estimations of total length according 
to the procedure employed may be relied upon as a criterion of assortive 
mating with respect to size. In figures 1 and 2, summarizing the ob- 
servations upon mating pairs, total length measurements are used. 
Figure 1 is a regression plot showing the correlation between the 
lengths of individuals and the average lengths of their mates, as found 
under natural circumstances. If the correlation were perfect, the ob- 
served points would lie upon line ''(1)"; if no correlation were to be 
detected the regression line "(2)" would coincide with m-m' \ actually 
the degree of correlation between the lengths of individuals composing 
copulating pairs is of about the same magnitude as that found in cul- 
tures of Paramecium containing a mixture of pure lines. 
Laboratory experiments were carried out with over 400 specimens, 
embracing about 200 individuals originally obtained in pairs, and 
an approximately equal number of ''single" nudibranchs. About 50 
specimens, ranging in length from 4 or 6 to 16 or 18 cm., were placed 
in each of a number of 9-gallon aquaria supplied with running water. 
After two days the mating pairs noted in each aquarium were removed, 
and measured. Data were in this way obtained from 119 pairs, which 
had come together under such purposely contrived circumstances that 
true random mating might easily take place, since so many individuals 
were crowded together in a small space; while if assortive mating is in 
any degree a real condition of copulation it should still make its infiu- 
