612 SCIENCE PROGRESS 



66 specimens, from 168 to 240 microns ; spine length, one 

 line of 23 specimens, from 9 to 46 microns ; another line of 1 16 

 specimens, from 46 to 149 microns. When measurements of 

 these characters were compared, it was found that correlations 

 often existed within the various lines ; for example, larger 

 specimens usually have a greater number of spines and teeth, 

 and larger spines than smaller members of the same line. 

 Different lines, however, differed in these correlations, indi- 

 cating that certain of the diverse hereditary characters vary 

 independently. The number of heritably diverse lines that 

 occur in nature is therefore very large. 



Selection was carried on with several lines ; the most 

 valuable results were obtained from one line which was con- 

 tinued over a period of 252 days, and contained 4,645 speci- 

 mens. The basis of selection used was chiefly past performance. 

 For example, when an attempt was made to obtain high and 

 low lines with respect to spine number, all members in one 

 line (the high line) were kept that had many spines, and that 

 produced offspring with many spines. If a specimen of the 

 high line that possessed many spines persisted in giving rise 

 to offspring with only a few spines, it was discarded. Similarly, 

 members of the low line were discarded if they produced 

 progeny with many spines, although they themselves possessed 

 none or only a few spines. Extensive experiments with this 

 large line resulted in the isolation by selection of heritably 

 diverse strains. The diversities occurred with respect to 

 number of spines, length of spines, number of teeth, and 

 diameter of shell, and since the last three named varied inde- 

 pendently, a large number of heritably diverse strains were 

 obtained (Fig. 4). While these heritable diversities apparently 

 were brought about gradually by small heritable variations, 

 larger changes were observed which might be brought under 

 the definition usually given to the word " mutation." For 

 example, in one line with an average diameter of about 190 

 microns, one specimen appeared that was about 40 microns 

 larger than the others ; the descendants of this specimen were 

 likewise about 40 microns larger than the other members of 

 the line. 



Root's (191 8) investigations on inheritance in asexual 

 reproduction were carried on with the fresh-water Rhizopod, 

 Centropyxis acaleata (Fig. 5), a species resembling Difflngia 

 corona in many respects. His work is not so extensive as 

 that of Jennings, but confirms certain results obtained by the 

 latter. It was first shown that diverse strains exist in nature. 

 The descendants of one " wild " specimen had on the average 

 about 6 spines, and divided on the average once every 6 days ; 

 whereas those of another " wild " specimen averaged 1-4 



