JANUARY 4, 1901.] 
charge had never been considered in connection 
with the atom save in relation to chemical and 
molecular effects. 
The last statement I shall criticise is the fol- 
lowing: He says (p. 888, 3d par.): 
“It is now pretty well established that the 
ether energy having to do with electrical at- 
traction and repulsion is dependent upon a sort 
of shearing distortion of the ether unaccompanied 
by any sensible diminution of volume, that this 
ether distortion is what is known as electric 
field, that the propagation of this energy con- 
stitutes electrical waves, and that the movement 
of the ether which comes into play during the es- 
tablishment of this shearing distortion, or which 
comes into play while distortion at one place 
is relieved and distortion at a contiguous place 
is built up, is what is known as magnetic field.”’ 
Surely not!! So far from being established, 
Dr. Franklin cannot adduce the slightest par- 
ticle of evidence for it. Though Maxwell and 
Lodge have used this theory, yet both Lord 
Kelvin and Professor J. J. Thomsun have sug- 
gested exactly the opposite theory, and Heavi- 
side has pointed out (Electromagnetic Theory, 
Vol. 1), that the theory which Dr. Franklin 
states is ‘pretty well established’ is at present 
as hard to reconcile with the facts as the other 
theory, so that the weight of authority would ap- 
pear to be fairly evenly divided. And one of our 
greatest living physicists, J. J. Thomson, uses 
the opposite theory, of late exclusively. More- 
over I have elsewhere pointed out that the varia- 
tion of / with the first power and of & with the 
second power is conclusive proof that the op- 
posite theory istrue.* If we chose to be uncon- 
* Those who are acquainted with my work on the 
nature of electricity and magnetism may remember 
that the proof that magnetism was a shear was based 
upon the following : 
(a) The determination of the fact that either & or 
# must be a density, thus confirming Williams’s 
result. 
(b) The demonstration of the fact that whichever 
one of the two & or # isa density, must depend upon 
the first power of the corresponding force, whilst the 
other must depend upon the second power of the cor- 
responding force. 
(ce) The experimental determination of the fact 
that / varies with H whilst k varies with F’. 
A second proof was then indicated, depending upon 
SCLENCE. 31 
vinced by this, then there is not the slightest 
evidence one way or another, and Dr. Franklin 
ean add considerably to his already brilliant 
reputation by producing some evidence in favor 
of his statement. 
REGINALD A. FESSENDEN. 
A BIBLIOGRAPHIC CATCH TITLE FOR THE YEARS 
1900 to 1999. 
In a note published in SciENcE, May 11th, I 
called attention to a bibliographic matter which 
I wish to return to again. 
Some twenty years ago I adopted the plan of 
placing all bibliographic titles at the end of an 
article in a single list with authors’ names ar- 
ranged alphabetically and each author’s papers 
arranged chronologically. Asan essential part 
of the plan, the citation in the text consisted 
simply in giving the author’s name and the last 
two figures of the year of publication preceded 
by an apostrophe. To avoid ambiguity, in case 
two or more cited papers were published by an 
author in one year, the abbreviated dates were 
followed by a lower-case letter used as an ex- 
ponent. This plan has been kept up since then 
in the ‘Contributions from the Zoological Lab- 
oratory at Harvard College.’ Owing to its 
simplicity and the evident advantage which it 
gives the reader by acquainting him at once 
with the date of the paper cited, this plan has 
come into rather common use. 
The apostrophe used to mark the omission of 
the first two figures of the year-date could not be 
used without ambiguity for dates subsequent to 
1899, and I have consequently urged in the note 
the nature of the Lagrangian terms involved in the 
change of & and y in elastic phenomena. 
I have now to add a third. Briefly stated it is 
as follows: Since either kor / is a density, then 
either H must be a shearing stress and Fa velocity or 
vice versa. It igs next shown that in the electric cur- 
rent we have a non-conservative system, and from 
quite general principles it is shown that it is the non- 
conservative system which must involve the veloci- 
ties. And it is shown that under no circumstances 
could the equation expressing the amount of the [?R 
loss be of the form it is if # were a shear, since in 
that case an operator which experiment shows is at- 
tached to an electric term would be connected with a 
magnetic term instead. 
