Mabch 27, 1903.] 



SCIENCE. 



495 



The Variations of Some Acquired Charac- 

 ters: R. P. BiGELOw and Eleanor P. 

 Rathbun. 



The discussion of the phenomena of co- 

 adaptation has emphasized the view that 

 success of the individual in the struggle 

 for existence may depend as much upon 

 favorable individually acquired modifica- 

 tions as upon congenital variations. The 

 present state of any species, then, is the 

 result of selection acting on both (1) varia- 

 tions and (2) modifications, tending to 

 eliminate the unfit of both indiscrimin- 

 ately, and to spare the best fitted. In order 

 to understand the effect of this process 

 upon the species as a whole it is necessary 

 to know the types and the distributions of 

 the deviations of the selected characters. 

 Much has been done to supply data in 

 regard to congenital variations by workers 

 following the methods perfected by Pear- 

 son. But heretofore this new means of 

 investigating biological phenomena has not 

 been employed in the study of acquired 

 modifications. The present investigation 

 was undertaken as a preliminary recon- 

 naissance of this new field. 



The material chosen for study is ob- 

 tained from the records of the first-year 

 students of the Boston Normal School 

 of Gymnastics, kindly furnished to the 

 authors by the director. Miss Amy M. 

 Homans. The records selected are those 

 of women who have completed the first 

 year of training and whose measurements 

 have been recorded at the beginning of 

 the year and at the end of the eight 

 months. The average age is 23.6 years, 

 but nearly half of the students are be- 

 tween 19 and 22 years of age. In most 

 of the series of measurements we were 

 able to obtain from 300 to 330 indi- 

 vidual records. The students come from 

 various parts of the country, and upon 

 entering the school are introduced into a 

 new environment, which is very uniform. 



the gymnastic and mental training being 

 the same for all students. 



The questions that the authors have 

 sought to answer are : (1) Is there a change 

 of type, and of what extent? (2) What 

 is the effect of training upon the variability 

 of the group? (3) What is the relation 

 between capacity for modification and ini- 

 tial position in the scale? and (4) what 

 relation exists between amoimt of modifi- 

 cation and length of time of training? 

 The first question has already been an- 

 swered partially by Beyer, Enebuske, and 

 Wood; the others have not been answered 

 before. 



Five series of measurements have been 

 studied, viz., (1) girth of left forearm, 

 (2) lung capacity, (3) mobility of chest, 

 i. e., difference between girth at rest and 

 girth at forced inspiration, (4) strength- 

 weight index, i. e., all the strength tests 

 added together and divided by the weight, 

 and (5) strength of legs. 



As was to be expected, the value of the 

 mean of each of these characters was found 

 to have become greater after training. 

 The difference is best expressed in terms 

 of the initial standard deviation. The 

 smallest change was in the girth of the left 

 forearm, amounting to 22.6 per cent, of the 

 standard deviation ; the greatest was in the 

 strength of legs, 162.5 per cent. The other 

 changes of type were: lung capacity, 40 

 per cent. ; mobility of chest, 55.5 per cent., 

 and strength-weight index, 101.2 per cent. 



The variability was found not to have 

 changed to a sensible degree in three of 

 the series, while in two others there was 

 an increase. For the strength-weight in- 

 dex this was 11.25 per cent, of the original 

 standard deviation, and for the strength 

 of legs, 18.33 per cent. It will be noticed 

 that these two series are the same ones in 

 which the increase of the mean exceeds the 

 magnitude of the standard deviation. The 

 frequency curves are all slightly skew at 



