626 
The entertainment and hospitality meted out to the 
Association by the Mayor and Corporation of Swansea 
and the leading citizens were of a most lavish and 
thoroughly enjoyable character, and the local arrange- 
ments were carried out by Mr. W. Grant Murray 
with a smoothness and precision which have never 
been surpassed. 
The Association, at its business meeting, elected a 
distinguished worker on the art side—Mr. E. Rim- 
bault Dibdin—to be president for 1914-15. 
THE HARDENING OF STEEL: 
‘AS the May meeting of the Iron and _ Steel 
Institute two papers were read and discussed 
dealing with the theory of the hardening of steel; 
they gave rise to one of the most interesting dis- 
cussions of the meeting. While it is, of course, 
obvious that the more ‘practical’? members of the 
institute take no interest in these discussions, it must 
be borne in mind that the theories of to-day become 
the foundations of the practice of to-morrow, and that 
therefore the ‘“‘practical’? man cannot in the least 
afford to despise or ignore what he likes to dismiss 
as ‘‘mere theory.” , 
The two papers on hardening presented to the insti- 
tute both put forward fairly definite theories, and at 
first sight these differ entirely from one another and 
still more from the older purely *‘‘allotropic’’ theory 
of hardening. It was interesting to find, however, 
that during the discussion not a single advocate of the 
other older theory—that of the so-called **carbonists”’ 
—came forward. When the views put forward by 
Profs. Edwards and Carpenter and by Mr. McCance 
are carefully compared, however, it will be found that 
they do not really differ very vitally either from one 
another or from the older allotropic view. All three 
theories agree in supposing that when carbon steel 
is cooled rapidly an essentially unstable transition 
product is formed which is itself intrinsically hard. 
The allotropic theory called this intermediate hard 
product ‘‘beta iron,’’ and identified it with the beta 
iron which has a limited range of stable existence in 
pure iron and in low-carbon steels; Messrss Edwards 
and Carpenter and the writer now identify it with the 
“hard amorphous phase” of Beilby, while Mr. 
McCance invents a new word and calls it “‘inter- 
strained”’ iron. The real difference of opinion seems 
to centre on the question how this intermediate sub- 
stance comes into existence. 
Profs. Edwards and Carpenter approach the subject 
from the point of view of an analogy between the 
hardening by quenching of steel and of alloys of 
copper containing from to to 13 per cent. of 
aluminium. These latter are somewhat hardened by 
quenching, and there is a corresponding similarity, on 
broad lines, between the respective constitutional dia- 
grams; by quenching, both kinds of alloys are caused 
to pass rapidly through a transformation range. The 
resulting micro-structures also show a certain simi- 
larity, the aluminium-copper alloys exhibiting an 
acicular structure having some resemblance to the 
coarser kinds of martensite seen in hardened steels. 
The authors then endeavour to show that the struc- 
tures of both quenched aluminium-copper alloys and 
of quenched steel arise from an identical cause, viz., 
highly multiplied twinning, which they believe to 
occur during quenching as a result of the internal 
strains caused during rapid cooling. The evidence 
that the martensite of steel is really equivalent to 
highly twinned austenite is, however, very weak, and 
it has yet to be proved or even shown to be likely 
that quenching can produce multiple twinning. In 
NO. 232721 OLEOs) 
NATURE 
[AUGUST 13, 1914 
the discussion of this paper, the writer pointed out 
that to produce notable strain-hardening of a plastic 
metal by deformation needed the application of really 
large deformations, while, on the contrary, the 
amount of deformation (strain in contradistinction to 
internal stress) which could be caused in steel by 
quenching must be very small. It is further a very 
open question whether strain really ever produces 
direct twinning in a metal. Finally, it has yet to be 
shown that a twinned constituent is really materially 
harder than in the untwinned condition; the softness 
and ductility of such materials as rolled and annealed 
copper or brass, which are one mass of twinned 
crystals, points in the opposite direction. The idea 
put forward by Edwards and Carpenter _ that 
amorphous layers are formed. on the twin boundaries 
in the process of twinning was well refuted by Mr. 
Humfrey in the discussion, as he showed by means of 
models that twinning could and did occur without dis- 
arrangement of the space-lattice at the boundary. As 
a result of the whole discussion it appeared that the 
authors had attached altogether too much weight to 
the process of twinning, but that the formation of 
amorphous metal during the quenching process might 
well be looked upon as the real cause of hardening— 
a view which the late F. Osmond had put forward 
quite clearly a year or so before his death. 
Mr. McCance’s paper began a thoughtful considera- 
tion of the whole subject by a review of existing 
theories. Both in regard to the existence or other- 
wise of beta iron as an independent allotropic modi- 
fication, and in regard to the amorphous theory, how- 
ever, the author made the mistake of considering 
that the objections which he raised can settle the 
point and dispose of these theories in half a dozen 
words. The question as to the extent and nature 
of the differences between alpha and beta iron is still 
being closely discussed, and even if Weiss’s magneton 
theory finds ultimate acceptance, it is still a question 
whether the magnetic transformations do not really 
constitute one type of allotropy, nor is it yet certain 
that they may not be associated with far-reaching 
changes in other properties. Again, as regards the 
amorphous theory, Mr. McCance’s objections are 
based on a simple misunderstanding, coupled with an 
assumption, based on the magneton theory, which has 
yet to be justified. This assumption amounts to the 
view that only crystalline solids can be ferro-magnetic, 
and that consequently amorphous iron would neces- 
sarily be non-magnetic. 
this view, for colloidal suspensions of iron are strongly 
ferro-magnetic, and so are a number of oxides and 
salts of iron, some of the latter even in a state of 
solution. 
Of much greater value are Mi. McCance’s experi- 
mental studies of hardening in carbon steel, which 
lead him, finally, to put forward a theory of the 
hardening of steel by ‘‘interstrain,’ which is prac- 
tically a translation of Osmond’s view ascribing the 
hardness of quenched steel to the presence in it of “le 
fer alpha écroui.”” It is only Mr. McCance’s account 
of the nature of “‘interstrained”’ iron which it is diffi- 
cult to accept. Declining to accept the views of 
Beilby and of Rosenhain as to the hardening of 
strained metals by the formation of layers of the 
amorphous phase, the author uses the word ‘ inter- 
strain” to denote a condition in which the regular 
crystalline arrangement is broken up generally, leav- 
ing a mass of irregularly arranged crystal fragments. 
It may well be asked what it is fhat holds these irre- 
gular and ill-fitting fragments together, and why an 
aggregate of such fragments should be harder than 
the aggregate of the larger pieces of crystal which 
constitute the ordinary soft metal? But beyond this 
Facts are, however, against. 
t 
