xlii Trans. Acad. Sci. of St. Louis. 



ago, all involved loads below the yield point of the concrete, which 

 had an age of seven years, instead of a month, which is the usual 

 age at which tests have been made. Repetition on the dry speci- 

 mens not only verified the previous conclusion as to its capacity to 

 safely endure an indefinite number of repetitions, but unexpectedly 

 indicate that repeated loadings did not lower the Modulus of 

 Elasticity materially below its early high value, which was perhaps 

 ten per cent greater than at an age of one month. When saturated 

 for a few days preceding the experiments, the Modulus had a value 

 of about sixty per cent of the value when dry, and repetition of load- 

 ing in this case also did not materially change its first value. 



Mr. Herbert A. Smith was elected to membership. 



OCTOBEK 21, 1912. 



President Engler in the chair ; attendance 32. 



Dr. Geo. 0. James addressed the Academy '*0n the 

 Contingence of the Physical Theory and the Problem of 

 the Geologic Past. ' ' 



After reviewing the theories of Helmholtz, Mach and Enriques on 

 the Principle of Causality, Dr. James stated that so far as descriptive 

 representation of the present is concerned, it makes no difference 

 whether or not we admit that pushing precision of measurement 

 further and further we shall ultimately come to a point where 

 there ceases to be accord between observation and theory, and be- 

 yond which it cannot be again established. Even adhering to this 

 conception of the contingent nature of the world of experience, we 

 may still assert that the results of observation agree with those of 

 theory to within errors which can be assigned. Experience alone, 

 however, does not justify us in concluding that successive approxi- 

 mations will make these outstanding errors as small as we please, 

 although this has long been the scientific conception. Admissible 

 or not, this conception must be referred to some postulate of knowl- 

 edge — the causal postulate as formulated by Enriques, for instance — 

 and becomes a metaphysical question. 



On the other hand, we may ask ourselves what we lose by denying 

 this exactness to the laws of nature, and reasoning on the basis of 

 the existence of an approximate physical theory only, which need 

 never become rigorously exact no matter how large be the circle of 

 phenomena taken into consideration, nor how precise the observa- 

 tions. 



Painlevfi's assumptions are: 



1st. It is possible to adopt for all time and for all phenomena 

 such a measure of length and such a measure of time that the 

 principle of causality will be true always and everywhere. 



2nd. It is possible to adopt for all time and for all measurements 

 of the universe, such a system of reference that the laws of mechanics 

 will be true always and everywhere. 



