498 TRANSACTIONS OF SECTION G. 
Engineers have not the same great difficulties which confront those who are 
advancing the boundaries of pure science; their models are very much what they 
please to make them; but, even then, problems arise which are sufficiently difficult 
to tax all the resources of applied science. The behaviour of models considered 
as similar structures is, therefore, a subject which engineers are bound to inves- 
tigate in order to determine the effects of fixed and moving loads, the action of 
wind, the pressure and frictional effects of steam and other fluids, and many 
other problems. 
In the majority of cases the simplest and the most direct method is the 
experimental study of a model, from which to obtain the data required for 
calculating effects on a full-sized structure, and hence the laws of similarity have 
received a very close scrutiny. 
Although most valuable information can be obtained from models, their use- 
fulness is clearly limited. The effects of the dead weight of a structure are pro- 
portional to the cube of the linear dimensions, and are, therefore, not usually 
measurable on a model except in exceptional circumstances, as, for instance, 
where elastic jellies are employed, as in the well-known investigations of Pearson 
on the stress distribution in reservoir dams. Nor are questions of stability easy 
to solve, since the forces producing instability are proportional to the size of the 
model. On the other hand, stress effects due to applications of load may be 
measured by the strains produced in a model of the same material, if the loads 
are proportional to the squares of the linear dimensions. The effects of applied 
load are studied even better in a model constructed of transparent material, since 
the variation of stress from point to point can be studied with much greater ease 
and certainty. 
As detailed models of this latter kind present some variations from the usual 
laws of similarity, it may be of interest to indicate their nature. Questions of 
deformation clearly involve the elastic constants of the transparent material and 
their relation to those of the proposed structure, while stress distribution in the 
solid is influenced by the value of Poisson’s ratio. This latter effect is quite 
small for glass, but may become appreciable with other substance. It is 
negligible in a model of any material which approximates to a thin plate stressed 
by forces in its own plane. 
The optical effects for any given load are, moreover, independent of the thick- 
ness of the material, and depend upon the stress difference, so that colour effects 
are obtained which may be regarded as pictures of shear stress throughout the 
model. Modern researches on ductile materials like structural steel indicate that 
such materials fail at some limiting value of shearing stress, and since the places 
where these limiting values are reached in the model are visible to the eye, the 
weak places in the design of a structure can be ascertained and a faulty design 
corrected by purely experimental means. 
In this connection it is of interest to mention that M. Mesnager, the chief 
engineer of bridges and roads to the French Government, has recently constructed 
an elaborate model in glass of a design for an arched bridge of about 310 feet 
span. This investigation was considered advisable for a work of this magnitude 
constructed of reinforced concrete, in order to check the calculations, especially 
of maximum stresses in the arched ribs, which iatter were assumed to be fixed at 
the ends. 
The effects of reinforcements were allowed for by determining equivalent 
sections of glass for the members of the model. Many difficulties had to be over- 
come in the production of a model free from optical defects, but these were all 
successfully surmounted. The stresses in the model were determined by aid of a 
Babinet compensator, and formed a valuable check upon the calculations for a 
structure of this great magnitude and somewhat unusual design. 
In this brief and incomplete account of a small branch of applied science 
relating to engineering the fundamental importance of discoveries in pure science 
is manifest. 
The discoveries in pure science and their innumerable applications to practical 
ends are ever a potent factor working for the common good, and the value which 
the British Association places upon applied science was most cordially voiced by 
Professor Bateson in his Portsmouth Address when he said : ‘ To the creation of 
applicable science the very highest gifts and training are well devoted,’ and, 
