modification, as embodied in instances of recent practice in which the cinder pot is dispensed with, will be illustrated hereafter.

An interesting contribution to the literature of the subject under consideration is to be found in the paper entitled, "Draft Appliances on Locomotives Exhibited at the World's Columbian Exposition," presented by Mr. Willis C. Squire, at the meeting of the Western Railway Club, November 21, 1893, and published in its Proceedings for that month (pp. 52-77). The ruling practice at that date is very

fully described and illustrated, and a basis of comparison with the examples of present practice, which will now be considered, is clearly outlined.

The recommendations of the 1894 report of the committee on 'Exhaust Nozzles and Steam Passages," met with general and continued favor, and have been adopted to a considerable extent, with satisfactory and approved results in service. The fact that the sixth conclusion, as to the shortening of the smoke box, has not been accorded with as promptly and generally as the other features of detail, is due largely to the fact that many of the long smoke boxes were made in sheets of a single length, which it was inconvenient to shorten, and that motive power officers were reluctant to incur the delay and expense of removing the long separate extension and substituting a shorter one. Force of habit and hesitancy to depart from the cherished traditions of the supposed advantages of the extended smoke box, were also, doubtless, factors of no inconsiderable weight. We therefore, find the construction of the committee adopted with both long and short smoke boxes, and now pass to the consideration of these and other instances of recent practice.

The application of the Master Mechanics' Association design in connection with an extended smoke box, is shown in Fig. 11, which illustrates the standard front end of the Michigan Central railroad. The extension of the deflecting plate in front of the exhaust pipe, before mentioned as having been discussed at the 1898 convention, as a means for clearing the smoke box of cinders, is employed in this construction, and the cinder pot or pocket is dispensed with, a plate being riveted over the opening, in cases where the cinder pot was formerly used. The utility of the extended smoke box in engines which are designed to, and in practice do clear themselves of cinders, is not apparent, and the reason for its employment is not stated in the letter of Mr. Robt. Miller, superintendent M. P. & E., M C. railroad, transmitting blue print of the front, dated March 31, 1899, in which he gives the following statement of his views and practice :

"I take pleasure in sending you blue print of the front end in use on the Michigan Central road. As stated in conversation there is no accumulation of cinders in the front end of Michigan Central engines equipped in this manner. It has not been the effort of this department to throw the cinders out of the stack, but to retain them in the firebox as much as possible. I do not believe a device was ever made that would entirely hold them. If there has been, I have not been fortunate enough to see it. However, I think we have been fairly successful in reducing it to the minimum. Experiment has taught me that a nozzle should be as large as the area of the exhaust opening, and never reduced below that point. Reducing the choke in the stack will have the effect of increased draft in the firebox the same as reducing the size of the nozzle, and at the same time will have the advantage of not causing back pressure. I believe I have followed the recommendations of the committee of 1896 of the Master Mechanics' Association on this subject, and have every reason to believe that by so doing I have accomplished as good results as could be obtained."

Mr. S. M. Vauclain, of the Baldwin Locomotive Works, has furnished ten blue prints illustrating different forms of front end constructions applied by his firm on engines built by them, regarding which he says: You will notice that some of these are so-called self-cleaning devices, or spark atomizers, and require

no cleaning at the end of a trip or for any number of trips. The dirt is thrown out but the fine particles only escape. I can give you no definite data on any of these systems. I am informed that each and every one of them is just right, and could send you more than a hundred others, also just right, so I decline to say which is the very 'rightest.'

Mr. Vauclain's blue prints all show extended smoke boxes, deflecting plates or diaphragms, and netting above and in front of exhaust pipe. Four of them substantially accord (except as to.length. of smoke box) with the Master Mechanics' 1894 design, and four are of the self-cleaning type, with forward extension of the deflecting plate. No features of special novelty or interest appear, and they are not therefore reproduced.

Fig. 12 shows the self-cleaning front end used on Class U engines, Norfolk & Western railway, regarding which Mr. W. H. Lewis, superintendent of motive power writes, under date of April 10, 1899, as follows:

“I take pleasure in enclosing you blue print mentioned, also revised blue print C 5262, showing practically the same arrangement with the low nozzle and double lifting pipes; both arrangements have proved entirely satisfactory, although in our service my preference is for the higher nozzle, as it affords a better opportunity to reach the flues in the front end, besides saving additional expense of the lifting pipes and their attachments.

"The self-cleaning feature of these are practically the same, and we have found that it is possible to do away entirely with the spark hopper and blower, and experience no trouble with the front filling with sparks.

"I beg to remind you, however, that the general rules to be observed in the arrangement of these spark-arresting devices are largely dependent upon the character of the coal and the service performed by locomotives, and the position of the deflector and diaphragm plates can only be determined by a careful service test The satisfactory results which we are now obtaining have thoroughly convinced us that it is not necessary to maintain a long extended front as a receptacle for sparks, and that only sufficient extension is required to insure the proper area of opening in the perforated plates or netting used to insure a free draught; in fact, in the number of observations which we made prior to the adoption of the device, it was found that our long extended fronts accumulated the maximum amount of sparks in a distance of ten miles, in heavy mountain service, which represented that all of the sparks which entered the front end after that time were necessarily thrown out. We, therefore, felt that little advantage might be expected from simply storing the sparks that would accumulate in going a distance of ten miles and running a further distance of fifty or sixty miles without any further accumulation, and that it was thoroughly logical to adopt a device which would relieve the front of sparks entirely."

Coming now to constructions in which a comparatively short smoke box, or one not longer, or materially longer, than is recommended by the Master Mechanics' Association in 1894, is employed, Fig. 13 shows the front end used on classes A, B and C engines of the Chicago, Burlington & Quincy railroad, all American type engines, with cylinders varying from 15 x 22 to 18 x 24, and may be taken as generally exemplifying the class in which a high exhaust nozzle is applied, without a petticoat or lift pipe. This arrangement, more or less modified in detail, is