444 JOURNAL OF THE WASHINGTON ACADEMY OF SCIENCES VOL. 13,,No. 20 
by Dr. Rosenuatn and his collegaues at the National Physical Laboratories. 
These included special microscopes, heating and quenching furnaces, special 
furnaces for the determination of thermal transformations and high tem- 
perature thermostats. The latter are used for the heating of alloy speci- 
mens for a period of several weeks or months at a constant temperature 
which has been proved necessary to obtain equilibrium conditions. The 
development of this thermostat has made possible the determination of the 
constitutional relation of certain alloys in which the phase changes take 
place very slowly and about which there has consequently been considerable 
doubt. 
Dr. RosEnHAIN also discussed briefly the organization of the department 
of metallurgy at the National Physical Laboratories and some of their prob- 
lems, particularly their work on the ternary diagrams of the light aluminum 
alloys. 
178TH MEETING 
The 178th meeting of the AcAprmy was held jointly with the Philo- 
sophical Society of Washington and the Chemical Society of Washington in 
the Auditorium of the Interior Building, the evening of Tuesday, April 17, 
1923. Dr. James C. Irvine, Principal and Vice-Chancellor, University 
of St. Andrews, and Dr. F. G. Donnan, Professor of Chemistry, University 
College, London, addressed the Societies concerning their own recent re- 
searches in chemistry. 
Principal IrviINE spoke on researches on The constitutional formula of 
starch. The first step was to determine the constitution of maltose, and 
this proved amenable to a method already developed by the speaker in 
which glucosides and complex carbohydrates are methylated before hydrol- 
ysis. Determination of the constitution of the scissive products thus ob- 
tained shows how the constituents of the complex molecule have been com- 
bined. By this method the maltose molecule was shown to be formed from 
two glucose residues by junction at the ends of the carbon chain. A similar 
process carried out on starch showed that two-thirds of the products cor- 
responded with those obtained from maltose, but the remaining one-third 
showed that a third molecule of glucose goes to make up the unit molecule of 
starch and that it is attached in the manner characteristic of cellulose. The 
simplest possible unit for starch is therefore one containing the nuclei of 
three molecules of glucose, two attached as in maltose and the third as 
mentioned above. An objection to this is that some experimenters have 
claimed a larger yield of maltose than the 72 per cent which would result 
from starch of this structure. These yields, however, refer to material 
which does not behave as a chemical individual towards bacteriological 
tests, and the most careful experimenters have declared that the maximum 
yield of the pure substance is about 70 per cent. 
Professor Donnan discussed Membrane equilibrium. He dealt with a 
number of cases of solutions separated by membranes which are imper- 
meable to certain of the ions. The exact equations for the relative concen- 
trations can be obtained by Gibbs’ method of equating the u-function for 
the ions which can pass the division. Approximate results may be ob- 
tained by using the usual dilute solution formula for the y-function. These 
approximations agree fairly closely with the results of a large number of 
experiments on solutions of various salts separated by ferrocyanide and 
other membranes. The predicted differences in e.m.f. across the membrane 
also agree fairly well with the facts. 
