166 MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 



anticipating an extended use of this system in the future, we have taken many measurements 

 according to this Agreement. 



Our decisions as to what to adopt and what to discard in different systems may appear 

 occasionally somewhat arbitrary; but they have usually been made in accordance with certain 

 rules which we have been constrained to adopt. We have not undertaken to sit in judgment on 

 the general relative merits of any system. All the systems extant are the results of more thought 

 and study than we have been able to devote to the subject of craniometry. We have accepted 

 that which seemed best suited to the scope of our work and to the character and extent of the 

 series to be studied. We have also had to take into consideration the limited time at our 

 disposal. 



Any measurement which we believed to be identical or practically the same in different sys- 

 tems we have taken but once and, in taking it, we have followed whichever rule seemed the most 

 explicit or laid down the most definite landmarks. Thus, in taking such a short dimension as the 

 interorbital width, where a small error may count for much, we have chosen for our landmark 

 the definite dacryon as directed by Broca, instead of the less certain ''inner border of the orbit," 

 which the Frankfort agreement prescribes for us. 



On some occasions we have discarded a dimension which had been made, or might be made, 

 the basis of extensive and valuable study, for the simple reason that we did not regard the given 

 directions as sufficiently explicit. While they might be clear to the scientist who wrote them, or 

 to the student who saw him apply them, they were not clear to the reader who had nothing but 

 the text to guide him. Thus we have taken no vertical measurements from the ophryou, because 

 no one tells us in what plane the connecting line between the frontal ridges should lie. Two or 

 more equally short lines between these ridges might, in some skulls, be described at some distance 

 from one another on the median line. In other words, we might have two ophryons so far apart 

 as to give materially different facial heights. We thought it better to be silent than to fill our 

 pages with material which might prove worthless. Had we had unlimited space, time, and assist- 

 ance we might have included such measurements, though not without comment. 



Some measurements were forbidden by the character of the skulls. The bistephanic diameter 

 was put aside because the temporal ridges in this series are indistinct, the indistinctness being 

 due, possibly, to the general use of boiled vegetable food among the Saladoans and the consequent 

 limited exercise of the temporal muscles. Moreover, the stephanion lies in a region especially 

 liable to be broken, and frequently was broken in the series. We have substituted for this dimen- 

 sion the maximum frontal diameter of Emil Schmidt.* The upper incisors were so often missing 

 that we neglected dimensions into which they entered. On the other hand, we took measurements 

 from the rnetopion, which is a very uncertain guiding point on these skulls in consequence of the 

 subdued character of the frontal bosses. 



We felt a great temptation to present to the reader such opinions concerning all the meas- 

 urements as we formed in the progress of our work, and to give our reasons in each case for 

 adopting this or abandoning that method; but on mature reflection we felt that this would lend 

 us beyond the proper scope of our work. In the more important cases, comments on the methods 

 are given in connection with the discussions of particular dimensions or indices. In some instances 

 we advanced far in the work of securing a dimension before we found practical reasons for aban- 

 doning it. In other cases we have taken a measurement on all the skulls of the series and com- 

 piled our tables and indices before we concluded to suppress our results. This we did, for instance, 

 in the case of the horizontal and vertical measurements of the orbit. 



We have, with some inconsistency, perhaps, adopted dimensions and followed rules of whose 

 exactness we felt no less uncertain than we felt of some which we discarded. Such instances are. 

 perhaps, to be classed among our arbitrary decisions. But we can partly atone for our errors, if 

 such exist, by telling exactly what our own methods were. For instance, we have recorded meas- 

 urements which have the superior border of the meatus auditorius for a guiding point, and we 

 must confess that we know not where to locate this point with accuracy. The rule for our own 



*EMII. SCHMIDT: Catalog <ler iin anatomischen Institut der Univereitat Leipzig aufgestelltou cramologiscken 

 Sammlung ; Archiv fiir Anthropologie, Braunschweig, 1887-'88, p. v. 



