SEPTEMBER 15, 1899. ] 
vestigator is much hampered by the diffi- 
culty of collecting all the references to the 
work in hand, scattered as they are through 
large and diffuse literatures; the univer- 
sity lecturer finds his utterances become 
more fragmentary every day; and as for 
the advanced student, he is often dismayed 
by the mere account of what must be done 
in order to be certain of the real state of 
opinion concerning the subject that interests 
him ; in fact,/we are likely to be crushed 
under the pressure of discoveries, old and 
new. The present Centralblitter and the 
Jahresberichten, excellent as they are, afford 
but scant relief. In every instance the 
seeker must examine the indexes of a long 
series of volumes, and these indexes are 
commonly the not too laborious creation of 
men uninstructed in cataloguing. Indexes, 
moreover, are usually put together from 
the titles of papers, and titles rarely give a 
sufficient idea of the the whole contents. 
But it is not my intention to write of the 
dark side of productiveness. The danger 
of swamping in the sea of literature is 
patent to every one. In several countries 
relief is being sought by the issue of card 
indexes. These remedial measures are 
alike in that they offer the titles of scien- 
tific communications, each printed on a 
single card, intended to be placed in an 
author’s catalogue, 7. e., a set of ‘cards ar- 
ranged alphabetically by the names. of 
authors. This function they might carry 
out very well, if they were printed and dis- 
tributed with sufficient promptness. 
Besides the author’s name, and the title, 
date, and place of publication, some addi- 
tional information usually is printed on the 
card so that duplicate cards may be ar- 
ranged as a subject catalogue, 7. ¢., filed al- 
phabetically according to subject. The ef- 
forts toward a subject catalogue, so far as I 
am acquainted with them, fall into three 
groups. In the first, it is proposed to give, 
besides the data already mentioned, a num- 
SCIENCE. 
ovl 
ber of cross-references, in other words, a 
list of matters treated by the author ; this, 
if I am correctly informed, is essentially 
the intention of the Royal Society. The 
second method proposes a few lines of text 
furnished by the author himself and stating 
his principal results; this was proposed by 
Professor H. P. Bowditch. The third con- 
sists of a few lines of contents, written by 
a cataloguer. 
Any of these catalogues is undoubtedly 
much more useful than a bare title. From 
none of the three, however, can the inves- 
tigator receive a satisfactory idea of the re- 
sults of his predecessors. The cross-refer- 
ence card is a simple index, and nothing 
more; the others, by reason of too great 
brevity, are not much better. Of the sec- 
ond and third method, I have some per- 
sonal experience. The American Journal of 
Physiology, at Dr. Bowditch’s suggestion, 
undertook a practical demonstration of his 
plan. An ‘index slip’ was issued with 
each number of the Journal. The slip con- 
tained the author’s name, date, title and 
place of publication of each article in the 
number, and a statement of results of not 
more than 150 words. 
Authors were invited to write their own 
statements. The slip was printed on thin 
paper, so that each statement could be cut 
out and pasted on a library card. In edit- 
ing the slip for the Journal, I found that 
the results of many investigations could not 
be stated in the space allowed; in such 
cases investigators objected with right that 
the too brief statement was misleading. It 
appeared further that nearly all the ‘ copy’ 
received from contributors had to be par- 
tially rewritten ; the author who had just 
filled many a broad octavo page could not 
shrink within the limits of the library card. 
A number of slips I had to write myself, 
because the authors failed to send any 
statement whatever, or sent them after the 
slip had gone to press. I believe that the 
