give plenty of room between panel points to contain all the necessary writing.

After the stress-diagrams are completed, the detail drawings are begun. There is considerable difference in the methods employed by consulting engineers to convey to manufacturers an understanding of the design which they desire to have executed in the shops. Some insist that the only proper method for the engineer to pursue, if he desires his details to be followed, is to make complete working or shop drawings, ready to be turned over to the template makers, while others prefer to make what are termed general detail drawings, which show to exact scale all the details, and give all important dimensions and the number of rivets in each connection, but which do not locate each rivet by figures, leaving the working drawings to be made by the manufacturer. When the latter method is adopted the working drawings must be sent in duplicate to the engineer for his approval before any of the work is sent into the shops, the said drawings being checked by the engineer's assistants, not only to see that they agree in every important particular with the original drawings, but also to make sure that they contain no errors of any kind.

The latter method is the one which the author invariably employs, and for adopting it he gives the following reasons:

First. Each bridge-shop has certain methods of doing work, which demand that the working drawings be made in accordance therewith; otherwise the cost of the manufacture is materially increased. These methods cannot be considered by the engineer, who has neither the time nor the inclination to go to the trouble of acquainting himself with the various methods of all the leading bridge-shops of the country.

Second. The nature of the work of a consulting engineer is not such as to justify him in keeping together enough trained draftsmen to execute with sufficient rapidity the large amount of drawing necessary, if the first-named method be followed.

Third. The capacity for accomplishing work in a consulting engineer's office when the second method is employed is probably three times as great as it would be were the first method adopted.