January 25, 1917] 
NATURE 
421 
‘never can be perfect, two ways of procedure are 
‘feasible : either to investigate the errors and disturbing 
factors and correct for them, or to select the condition 
‘of experiment so that the disturbing factors are 
negligible—for instance, experiment on a large scale. 
The latter method cannot give as high accuracy as 
‘the former, but the former method, while theoretically 
more accurate, may give a constant error, possibly of 
hundreds per cent., if some of the assumptions on 
which the corrections are based are not completely 
-justified. Industrial research leans towards the first 
method as giving results which are safer in trustworthi- 
ness, even if somewhat less accurate, while educa- 
tional research leans towards the method of applying 
corrections. As illustration, in magnetic investigations, 
«the effect of joints in the magnetic circuit, etc., 
may be determined and corrections for it applied, or 
such a magnetic circuit may be chosen, that the effect 
of joints, etc., is negligible, and can be neglected, or 
taken care of, by a correction which is so small that 
its accuracy is not material. 
In industrial research the liability exists of limiting 
the work to such a narrow field that it has little 
general scientific value; for instance, to detérmine the 
hysteresis loss in a magnetic material, without deter- 
mining the magnetisation curve. In educational -re- 
search inversely there is sometimes the tendency to 
generalise beyond the limits justified, and so draw 
wrong conclusions. For instance, numerous investi- 
gations have been made and conclusions drawn there- 
from in treatises on the ‘“‘arc,’’ while in reality the 
investigation. was made with the carbon arc only and 
applies only to this kind of arc; and as ‘the carbon 
arc is not typical, but rather exceptional, for most other 
‘arcs the conclusions are wrong. 
As regards the quality of the scientific research work 
done in industrial organisations compared with that 
done in educational establishments, there is no material 
difference, but the work done in ‘the industry, just as 
that done in universities, varies from scientific research 
of the highest quality down to investigations which 
are of little, if any, value—investigations crude and in- 
accurate or directly erroneous in premises, in method, 
and in results and their interpretation, or investiga- 
tions which, while correctly conceived and correctly 
‘made, are useless because essential conditions have 
not been controlled or recorded. Still worse are those 
pseudo-scientific investigations occasionally met which 
owe their conception to the desire of self-advertisement 
or are made for commercial or legal purposes, such 
as, for instance, to give the appearance of a scientific 
standing to some theory which some inventor had 
recorded in his patents. Such work—met occasionally, 
though less and less frequently—in industrial as well 
as in educational institutions, tends to discredit scien- 
tific research in the eyes of the layman, who cannot 
discriminate between science and pseudo-science. 
The essential difference between industrial and edu- 
cational research, however, is met in their method of 
publication: the publication mediums of scientific re- 
search carried on in educational institutions are the 
scientific publications published more or less under 
the direction or supervision of universities, while the 
publication mediums of the scientific research carried 
on in the industry are the technical or engineering 
‘papers, and only occasionally an abstract reaches the 
scientific publications. Unfortunately, a large number 
of men of science still look on publications in the tech- 
nical Press as unscientific, take no cognisance of them, 
do not recognise them in scientific abstracts, reviews, 
etc., and, as a result, a large and steadily increasing 
part of the scientific research of the country is prac- 
tically lost to men of science, and is not available or 
easily accessible, by not being recorded, abstracted, or 
NO. 2465, VOL. 98] 
indexed in the records of scientific progress. If, for 
instance, in the tables of physical constants published 
only a few years ago, under “‘hysteresis’’ are pub- 
lished the losses in a Siemens cable transformer (a type 
which had ceased to exist a quarter of a century ago), 
and practically all the mass of data on magnetism 
recorded in the engineering proceedings neglected, 
apparently as not ‘‘scientific,” it shows that there is 
something wrong with the attitude of those respon- 
sible for the records of science. Amongst the worst 
offenders in this unjustified exclusiveness are the 
physicists, while the chemists make a commendable 
exception. In the “chemical abstracts” published by 
the American Chemical Society, the results of indus- 
trial research, as well as those of the chemical univer- 
sity laboratories, are recognised, and these abstracts 
are therefore comprehensive and valuable, which 
cannot be said of the abstracts of some other sciences. 
Possibly the reason is because applied chemistry is 
chemistry just as well as theoretical chemistry, while 
applied physics goes under the name of engineering, 
and the average theoretical physicist is rather inclined 
not to recognise engineering as scientific. : 
Some excuse may be found in the nature of the two 
classes of publications, the physical science publications 
and the engineering publications. The former accept 
for publication only scientific papers, exert a critical 
judgment, and the appearance in the scientific publica- 
tion medium ‘thus implies that the article, at. least in 
the opinion of the editors, is of scientific value. This 
is not the case, and cannot be the case, with the 
engineering or technical publications. The technical 
Press is the medium of all the publications of those 
engaged in the industry, from scientific research of 
the highest value to mere commercial statements, and 
the appearance of an article in an engineering paper 
or transaction does not imply, nor is intended to imply, 
that it is of scientific value, but the discrimination of 
scientific worth, which in the scientific publica- 
tions is attempted by the editors, has in the 
engineering Press to be left to the reader or abstractor. 
If, however, the purpose of the engineering publica- 
tion is to bring together all classes of industrial 
records—and it thus includes commercial and other 
articles—this is no justification to refuse recognition 
to scientific papers contained in the same publication, 
but rather makes it desirable, and indeed necessary, in 
the interest of our nation’s scientific efficiency, to find 
some means or organisation to carry out this dis- 
crimination and make available to the scientific world 
at large the scientific work contained in the annals 
of applied science—that is, engineering, 
UNIVERSITY AND EDUCATIONAL 
INTELLIGENCE. 
Oxrorp.—In the course of the term just beginning, 
Congregation will be invited to consider an important 
proposal for establishing a new status for advanced 
students, and for enabling persons admitted to this 
status to obtain the degree of doctor under new condi- 
tions. It is proposed to limit admission to the status 
to persons who have taken the degree of B.A. at Ox- 
ford; or who, if they come from another university, 
have taken a four years’ course and a degree, and 
have produced satisfactory evidence of their fitness to 
pursue a course of advanced study. The time to be 
devoted to study by advanced students before the degree 
of doctor can be taken will be either two or three years, 
according to circumstances. Opposition may be ex- 
pected to this proposal, both in principle and detail, 
and it is yet too early to forecast the result of dis- 
cussion. The advocates of the scheme spealx of it as 
an attempt to meet a need which is likely ‘to be. of 
: considerable importance at fhe end of the war. 
