420 
MR. F. OSMOND AND PROFESSOR ROBERTS-AUSTEN 
reason to think that many disturbing causes are, if not eliminated, at least reduced 
to a minimum in this series of alloys. It will be possible up to a certain point to 
apportion the effect of these disturbing causes, and eventually to set aside such 
complications as tend to conceal the effect of atomic volume in the researches to 
which reference has been made. 
With the exception of some facts stated by Behrens, and of some early experi¬ 
ments of our own, on the alloys of gold and aluminium, to which we have just 
referred, Professor Arnold was the first to examine the alloys of gold. He took 
the experiments of one of us on the influence of impurities on the mechanical 
properties of gold as the basis of his work,* and made some careful drawings of 
etched sections of pure gold alloyed with about 0‘2 per cent, of various impurities. 
We consider, however, that photographs of micro-sections are far preferable to 
drawings, and we agree with M. Cbarpy,+ in thinking that drawings, however careful, 
give a very incomplete idea of the appearance of etched metallic surfaces. We have 
therefore great pleasure in submitting to the Society micro-photographs from the 
identical specimens of gold described by one of us in 1888, which formed the starting 
point for the later investigations. The results of the micrographic examination of these 
specimens of gold forms the first part of the present paper. 
Part I. 
The alloys examined may be conveniently represented by the chemical symbol of 
the metal added to the gold, which was in each case in the proportion of 0'2 per cent. 
They are the twelve following : K, Bi, Zr, Bh, In, Li, Se, Zn, Pd, Te, Sb, Ah Thus 
“ K” would signify gold containing 0'2 per cent, of potassium and so on. It will 
also be convenient to supplement this series by a sample of pure gold which was 
cooled under precisely the same conditions as the “ test pieces ” into which the alloys 
submitted to mechanical tests were cast. 
These test pieces had been divided by two chisel cuts close to the part which 
had been restricted by traction, so that the portions which were preserved resembled 
the form shown in fig. 1 . The structure would have been examined on a section at 
Fio . j right angles to the axis of the bar, but this method of procedure 
proved to be inconvenient, as it did not make the most of the 
precious metal, and it would have notably diminished the dimen¬ 
sions of the fragments, which were already too small to be con¬ 
veniently handled. It was necessary, therefore, to be content 
with rubbing the face a b on emery paper until it became flat, 
hence the sections are oblique in relation to the axis of the bar. W T e should have 
preferred to examine bars which had not been distorted by traction, but it was 
important to use the old specimens which were available. 
* ‘ Engineering,’ vol. 61, 1896, p. 176. 
f 1 Bull. Soo, d’Encourage,’ vol. 1, 1896, p. 200. 
