G.— ENGINEERING. 149 



a motor-car along a road. Cam shafts for the stamping mills, used in the 

 gold mines in the neighbourhood where we are, frequently fail, and 

 engineers and metallurgists are often at a loss to give a reasonable expla- 

 nation. The whole problem is bound up, not only with the physical pro- 

 perties which statical experiments are capable of testing, but apparently 

 in some unknown way with the nature of the crystal structure of a metal ; 

 perhaps the stress that produces fracture depends upon the distribution of 

 very small quantities of so-called impurities in the space lattice of the 

 crystal, or upon the nature and properties of the material forming the 

 crystal boundaries, and upon the surface conditions, including hardly per- 

 ceptible corrosion, as well as the existence of other discontinuities incidental 

 to the manufacture of metals. Some metals appear to fail by slipping on 

 the crystal plane, and others by cracks commencing at crystal boundaries, 

 or at surface discontinuities. To the metallurgists, engineers must 

 unstintingly pay tribute for the splendid work they have done in supplying 

 materials to meet the ever growing and exacting needs of modern engineer- 

 ing. By the careful scientific control which characterises a modern steel- 

 works, thousands of tons of apparently uniform material can be turned out, 

 but the engineer does not always use the materials as they should be used, 

 and the metallurgist has not always the knowledge required to insure 

 perfect reliability under the exacting conditions of service. These, allied 

 to the difficulties to which reference has been made, make it imperative 

 that in this problem of the reasons for failure of metals under repeated 

 stresses and their prevention, the co-operation of engineers, metallurgists, 

 physicists, and mathematicians must be enlisted ; experience alone is by 

 no means satisfactory. 



Incidental to the failure of materials in service, the recognition of 

 synchronism as an important contributing factor, and the assistance given 

 by experiment and dynamical theory in anticipating synchronous speeds, 

 has been of the greatest service in the design of modern high-speed machines. 

 Crank shafts of aeroplane engines and of internal combustion engines 

 have been known to fail in a few minutes because designers have failed to 

 recognise the possibilities of synchronism.* Such disasters can now be 

 avoided in many cases. 



Roads. 



Before concluding, I should like to refer to another subject of interest. 

 The rapid developments of road traffic during the last quarter of a century 

 has made the road problem one of considerable importance. From the 

 point of view of the user and the community, the general policy of road 

 construction, the capital expenditure involved, and the cost of upkeep are 

 of primary concern. These are dependent upon the materials available 

 and the use made of them. Incidental to the problem is that of vibrations 

 produced in structures adjoining the road and the probable damage 

 accruing from them, but of this there is only time to say that experiment 

 is necessary to determine the importance of these vibrations on the 

 materials of structures. 



Reference has already been made to the lack of knowledge of the 

 effect of blows and stresses due to vibrations in causing failure in metals, 

 and the lack of precise knowledge which will make it possible to choose with 

 assurance the particular material to meet certain circumstances. The 



