188 



SCIENCE. 



[N. R. Vol. VII. No. 163. 



A single illustration further will, per- 

 haps, suffice on this head. A few years ago 

 a shipment of some three hundred freight 

 axles was received at two diiierent shops on 

 the Pennsylvania Railroad, from an entirely 

 reputable maker. Some of these axles were 

 used for repairs, and some went under new 

 cars. Scarcely had they gotten into service, 

 however, before difficulty began to arise. 

 The axles began to break. Indeed, one of 

 them broke before the car had been turned 

 out of the shop j'ard, one broke into thi'ee 

 pieces before the car had made 150 miles, 

 and in less than three months eight had 

 broken. Each of the broken axles was sent 

 to the laboratory, and a careful study of 

 the case made, with the hope of discovering 

 the cause of the failure. An examination 

 of the freshly fractured ends of several of 

 the broken axles showed that for a little dis- 

 tance in from the circumference the frac- 

 tured steel presented an appearance quite 

 different from that given by the remainder 

 of the fracture. Moreover, a line of de- 

 marcation between these two apparently 

 different kinds of steel in the same axle 

 could be clearly traced. Accordingly, it 

 was decided to make analysis of borings 

 from near the circumference and near the 

 center, and see whether this would reveal 

 anything. It may be stated that the axles 

 were known to have been made from Besse- 

 mer steel, and should normally have con- 

 tained not more than 0.10 per cent, of 

 phosphorus. The analysis of the borings 

 from near the circumference of the axles in 

 no case gave figures up to this limit, while 

 the borings from the center of the axles in 

 no case showed less than 0.16 per cent, 

 phosphorus, and in some cases the amount 

 was as high as 0.24 per cent. Those who 

 are familiar with the methods in daily use 

 in modern steel works will, from these fig- 

 ures, at once understand the cause of break- 

 age of these axles. For the benefit of those 

 who are not, it may be weU to explain that 



in most modern steel works large ingots are 

 now the rule, and that in large ingots, 

 which take considerable time to solidify 

 from the molten condition, analyses show 

 that some of the constituents of the steel 

 are not uniformly disseminated throughout 

 the mass. This separation of the constitu- 

 ents during cooling, technically known as 

 ' segregation,' is characteristic of the carbon, 

 the phosphorus and the sulphur. Further- 

 more, the segregation appears to be worst in 

 the upper third of the ingot, so much so 

 that many specifications now require the 

 upper third of the ingot to be removed, and 

 not used at all in making the articles the 

 specifications call for. This much being 

 stated, it is clear why our axles broke. They 

 were made from badly segregated steel, per- 

 haps from the rejected upper thirds of a lot 

 of ingots, the balance of which were used 

 for other purposes. Subsequent correspond- 

 ence with the parties furnishing the axles 

 gave good grounds for belief that such was 

 the case. For the comfort of those who ride 

 on railroads, it may be added that the 300 

 axles were at once withdi-awn from service, 

 and that since that time a chemical and 

 physical specification for both passenger and 

 freight axles has been prepared which is be- 

 lieved to preclude the possibility of such 

 axles as are described above being received 

 by the Pennsylvania Railroad. 



These illustrations of the power of an 

 analysis to explain difficulties could be pro- 

 longed to almost any extent, but I spare 

 you. Furthermore, I should not like to be 

 understood as claiming that every puzzle, 

 every difficulty or every state of affairs in 

 nature where the reasons for the phe- 

 nomena which we find are not apparent 

 at sight can be explained by a chemical 

 analysis. Our knowledge is far too limited 

 for this. Moreover, many cases could be 

 cited in which an analysis throws no light 

 whatever on the situation; but, notwith- 

 standing this, an experience of some twenty 



