500 THE AMERICAN NATURALIST. [Vor. XXXV. 
more errors necessary to obviate, more diversity of the manner of 
action, than dissection. ‘Far from being more elementary, anthro- 
pometric researches presume a preceding acquisition of knowledge 
by dissection." 
Anthropometry is “a procedure of anatomical analysis, serving to - 
make more precise the description of the innumerable variations of 
the human body." A thorough knowledge of all these variations, 
unattainable by unaided observation, is necessary to anthropology, 
which, in the definition of the author, is *a synthetic knowledge of 
the human beings, the species, groups, categories, individuals." In 
order that anthropometry should prove of full value and its data lend 
themselves with the greatest attainable facility to biological interpre- 
tations, it requires in practice the greatest possible precision and the 
maximum of rationality from the anatomical standpoint. 
In all his work the one who measures should have a well-defined 
aim, but be free of preconceived notions ; he should have a precise 
definition as to what is to be measured and registered, at least an 
approximative estimation of the aptness of the instruments used to 
answer the requirements, and a practical knowledge of the technique 
of his measuring; he must be able to exercise a proper individual 
discretion in connection with the various irregularities encountered, 
and capable of a correct biological interpretation of the conditions 
met and data secured. Such competencies, it is self-evident, require 
a long preliminary preparation in anatomy, physiology, and other 
branches of science, as well as a practical course in measuring. 
The best preparation for anthropometrical researches is found in 
the ensemble of medical studies; even these studies, however, 
are not sufficient until supplemented by a special anthropometric 
preparation. 
A text-book instruction in anthropometry is not sufficient ; it is 
the practical apprenticeship in the operatory technique of anthro- 
pometry which becomes more and more indispensable; this fact 
commences to be generally acknowledged. 
Unnecessary modifications of simple and good instruments or 
methods are pernicious. It is possible to conserve and progress at 
the same time. 
It is most important that ciphers obtained by one investigator 
should be comparable with those of others ; if there be any personal 
errors resulting from the technique or other causes, there results 
only an apparent comparability, which is false in some particulars 
and leads to more, or less considerable errors. 
