table indicates the difficulty in comparing coal records, but the conclusions would be that the 700 and 710 classes are equal to the 1300 class between Winnipeg and Kenora, slightly inferior to it between Kenora and Ignace, and slightly superior between Ignace and Fort William, while on the latter section the 740 class are decidedly more economical than either of the others. It is evident, however, that to average these figures would not give an accurate result on account of the variable amount of work done each month. The most satisfactory plan, if records were absolutely accurate, would be to continually compare with one class and the average increase or decrease could then be obtained. It can easily be seen that if the class compared with, used a small amount of coal during any one month the comparisons made with it might be widely misleading; for instance were the 1300 class taken as a standard, the results of the other engines would show an enormous economy in January between Ignace and Kenora. If, however, in each month the engine using the largest amount of coal is taken as a standard, and the amounts which the other classes would have consumed at its rate are calculated, a comparison of the coal actually used as against that which would have been used, is obtained, which affords a reasonable measure of their economy. A hypothetical illustration of this method is as follows:

This method tends to minimise the variations which occur and is thus conservative in its results; it is also inapplicable where the classes of en

gines to be compared do not work together on the same sections to a sufficient extent to obtain fair comparison, and is thus in some ways not as satisfactory as deductions drawn from a study of the records, but it has the advantage of eliminating all personal element and is the only system so far devised by which any reasonable compilation of records extending through several months on various sections can be made.

The records given above for the Central Division, September to January inclusive, when compared in this manner yield the following results:

The records on the Lake Superior Division for the months of May to January inclusive when compiled in the same way give the following results:

It is interesting to note that the above results may also be compared as follows using the compounds of each type as a basis.

Assuming these comparisons to be accurate and after careful study the writer certainly believes them to be conservative, the superheater loco

motive has evidently shown itself to be rather more economical than the compound with the amount of supperheat obtained; on the "Schmidt" and "Schenectady B" design, and from the good results obtained from the

FIG. 8-BOTTOM OR SUPERHEATED STEAM HEADER.

C. P. R. engines there is every reason to believe that still further economies will be reached. The results from these engines were entirely unexpected as the design was developed from a mechanical standpoint entirely,

and they are not by any means easy to explain, except that they must be caused by the steam being heated to a higher temperature than in the "Schmidt" or "Schenectady" engines. Since this was not discovered until

December, it has been impossible to carry out proper temperature tests on both types, but one made on one of the 740 class engines would indicate a higher superheat of about 40°. In view of the superheating surface be

ing identical, this can only be explained by the peculiar arrangement of the headers which prevents any abstraction of heat from the steam after it has been superheated by the entering steam, by the tendency to a more

UNIVERSITY OF WISCONSIN.

COLLEGE OF MECHANICS & ENGINEERING

EUGENE DIETZGEN CO. CHICAGO.

equal flow of the steam through the various superheater pipes in passing from one header to another. In support of the first reason the poor results obtained from the "Schenectady" A type are of interest. With 340

square feet of superheating surface against 375 in the "Schmidt" engine, the superheat was only about 20° as against 100°. This would show that

COLLEGE OF

CHANICS & ENGINEERING

EUGENE DIETZGEN CO., CHICAGO.