240 
MCGEE. 
which they draw at will in making up each issue; in the 
literary and popular magazines some articles are so kept on 
hand for months and sometimes years; and even in scien¬ 
tific journals of the highest grade individual memoirs are 
often held for weeks or months before they can be fitted into 
the Procrustean bed provided for each family. Moreover, 
editorial labor is required in rounding out the various de¬ 
partments into which the periodical is divided, and this in¬ 
volves added expense. These impediments of delay and 
cost appear to be the rocks upon which the periodicals 
launched by most of the societies have been wrecked. 
The author pursues his researches con amove, and on pub¬ 
lication of results finds among his fellow-students a large 
circle of readers who share interest in his problems. So the 
inspiration of authors and the enthusiasm of students breed 
special treatises; and it is the chief function of the scientific 
society to bring the author into communication with the 
student and to aid the labors of the former and supply the 
needs of the latter. But except in rare cases there is none 
of the inspiration of authorship in the preparation of admin¬ 
istrative records of scientific societies, and there are few care¬ 
ful readers of such matter save among historians or pro¬ 
jectors of new scientific institutions. Such records are in¬ 
deed necessary for the society and in a less degree for the 
public, but it is evident a 'priori that the incentives to prep¬ 
aration and the demand for distribution and reading of 
“ proceedings ” must be limited, and that such records must 
ever be parasitic upon the scientific records which it is the 
special province of the society to maintain. Here, again, de¬ 
duction from the fundamental principles controlling publi¬ 
cation, and induction from the history of scientific serials 
are coincident; for although nearly all the societies have 
sought to support separate administrative records, only two 
are now doing so, and in one of these the serial is unstable 
while in the second the conditions are unique. 
In brief, it appears that the variations and tendencies to 
vary exhibited by the serial publications of scientific socie- 
