230 



SCIENCE. 



[N. S. Vol. I. No. 9. 



be in all his measurements on the same per- 

 centile grade. This latter method is based 

 on a quite erroneous theory of the propor- 

 tions of the body. Dr. Porter's method is also 

 better than that based on single measure- 

 ments, as it points out abnormal propor- 

 tions, not simply abnormal size. It is 

 necessarj^, however, to bear in mind 

 the one restriction that many measure- 

 ments are not closely correlated with 

 stature, but have different correlations. 

 This is the case with girth of chest, strength 

 of squeeze and many others. Therefore 

 their correlation to stature will not give 

 more satisfactory results than the study of 

 the single measurements alone. It will 

 certainly be of great use to school hygiene 

 to subject all children whose proportions 

 are abnormal to a medical examination, 

 but it will not be possible to determine by 

 means of the measurements what indi- 

 viduals are retarded in growth and what 

 are advanced, as Dr. Porter suggests, ex- 

 cept in very exceptional cases. The corre- 

 lation between any two measurements is so 

 slight that a great many cases which are 

 noi'mal for one year are also quite normal 

 for the preceding and following years at 

 least. This is also shown by the fact which 

 is apparently so contradictory, that children 

 of a certain height are the heavier the older 

 they are (according to Bowditch), but that 

 also children of a certain weight are the 

 taller the older they are. 



Finally, I must say a word in regard to 

 Dr. Porter's objection to the combination of 

 measurements taken in different cities. It 

 is, of course, true that the results in various 

 cities depend upon the composition of the 

 population and its geographical and social 

 sxirroundings. If we knew all these factors 

 and their influences it would be necessary 

 to sub-divide the series of each city into 

 numerous divisions. As we do not know 

 the exact influence of these factors, we must 

 endeavor to take as our basis a general 



curve, including as many individuals as 

 possible of the same population but under 

 a diversitj^ of conditions and compare the 

 curves determined by certain factors with 

 them. It is, therefore, perfectly correct" 

 to compute the growth of American chil- 

 dren from data collected in various cities, 

 provided each city is given its proper 

 weight according to the number of chil- 

 dren measured. The more cities and vil- 

 lages are included in such a combination, 

 the more nearly we shall get the curve re- 

 presenting the growth of the American child. 

 By comparing the general curve with the 

 ones obtained in different cities we can in- 

 vestigate the causes which produce the dif- 

 ference between the individual curves and 

 the general curve. We know that national- 

 ity, occupation, social status have a con- 

 siderable influence. I have found that first- 

 born children exceed later-born children in 

 size. The effect of all these causes can be 

 studied bj^ comparing the individuals repre- 

 senting each group of factors wth the gen- 

 eral population. Feanz Boas. 

 New Yoek. 



LABORATORY TEACHING OF LARGE CLASSES 

 IN BOTANY.* 



The great increase in the size of the classes 

 in Elementary Botany during late years in 

 Harvard College has forced their teachers 

 to the development of some system for 

 their efficient and economical management 

 in the Laboratory. Under the guidance of 

 Professor Goodale there has been worked 

 out the plan upon which are based the 

 recommendations made in this paper; in- 

 deed, what I have to say is little more than a 

 description of the system in use there during 

 the last year I was connected with it, i. e., 

 1892-'93. My observations are, therefore, 

 based not upon theory alone, but upon the 

 results of trial and selection. 



*Eead before the American Society of Naturalists, 

 Baltimore, Dec. 38, 1894. 



