382 
INTERNATIONAL ASTRONOMICAL UNION Proc. N. A. S. 
bility of a central committee charged with decisions relating to notation, 
nomenclature, etc. For example, in the matter of naming newly dis- 
covered variable stars, endless confusion can be avoided only through the 
action of a centralized body of recognized authority and responsibility. 
Nevertheless some elaboration of statement is necessary in order to em- 
phasize what we believe to be a sound basis of procedure. 
A sharp distinction should be drawn between the adoption, on the one 
hand, of definitions and conventions, and, on the other, of the numerical 
results which flow from the introduction and usage of such definitions and 
conventions. The adoption of a definition or convention naturally should 
never be undertaken hastily or without detailed consideration of all the 
attendant circumstances; but once an appropriate formulation has been 
attained, there need be no hesitation or objection to its universal acceptance. 
Thus we find ever5rwhere in use the definition of the scale of magnitudes 
represented by the formula 
m ~ nio = -0.4 (log I - loglo). 
Similarly, the convention adopted by the committee on magnitudes of the 
Carte du del relative to the zero point of the photographic scale of mag- 
nitudes finds ready acceptance on the part of all observers. Or again, the 
definition of color-index as the difference: photographic magnitude 
minus visual or photovisual magnitude, affords so convenient and useful 
a measure of the color of a star that it meets with the approval of every- 
one engaged with the problems of stellar photometry. Thus we find 
introduced into astronomical practice a uniformity and standardization 
which is only useful and helpful. 
The practical application of these definitions and conventions requires, 
however, the usage of numerical standards and constants. And here 
caution must be exercised if progress is really to be aided. A definition 
or convention in the nature of the case possesses elements of permanence 
and invariability ; but a numerical constant is a thing of the day; to be dis- 
carded the moment a more precise determination becomes available. The 
quality of inconstancy, the ever-increasing precision which characterizes 
the metrical results of science, should always be borne in mind; otherwise 
there is danger that the seal of authority may be placed upon results in 
which authority should have no place. The thing to be emphasized is 
evidence, not authority, and the evidence should be frequently reviewed. 
Thus in the matter of magnitudes: it will be necessary in undertaking 
any cooperative investigation to specify the standards which are to be used. 
The selection naturally will be made by the International Committee 
after a critical examination of the evidence underlying the results then 
available; but it should be clearly understood that the standards thus 
chosen are for that particular investigation and are not necessarily to be 
used for any other which subsequently may be undertaken. In the present 
