The following text may have been generated by Optical Character Recognition, with varying degrees of accuracy. Reader beware!

26 fact that not sufficient importance was attached to the riveting of the webs to the shell, and, where double bottoms were fitted, the connection at the margin plate left much to be desired. On the whole THE MarRINE. REVIEW to the shell, to support the deep frames at intervals between the deck and floor (see Fig. 2). It eventually became obvious that this was a great waste of material, oo a a a @ L LO FIG. 3--LENGTH BETWEEN PERPENDICULARS, 346 FT. 6 IN.; BREADTH, EXTREME, 50 FT. 8 IN.; DEPTH, MOLDED, 25 FT. 6 IN. we must admit that it was not sufficiently well understood that thé webs, acting in conjunction with the side stringers, had to be strong enough to support all the intermediate frames. The expensive form of the riveted attachments necessary for web framing encouraged the development of the deep frame system, obviously a more scientific arrangement of material than that adopted in web-framed vessels, inasmuch as a proper frame girder is se- cured, and each frame in this system is able to take care of its own portion of transverse strength without borrowed support, with the exception of that ob- tained from the shell plating as previously explained. The increased demand for single deckers of larger size quickly 'brought up the question of compensating for the omission of all decks, except the upper deck, and a succession of rather extraordinary types of framing was the result, the most curious being a mixture of the web-frame system and the deep- frame system, with a tier of beams at the height of the lower deck (see Fig. 1). It is obvious, in this system,.that if the web frames were of any use at all in supporting the intermediate frames, there was no reason whatever for fitting these frames according to the deep frame rule. Further, if the web frames were not strong enough to support the intermedi- ate frames, there was no object whatever gained in fitting them, as deep frames of the same size as the other deep frames would have been efficient. Turning back to the deep-frame system without web frames, I need not remind my hearers that a strong attempt was made by fitting H or double bulb angle stringers, with an intercostal plate riveted usually so great that these stringers were ineffectual as girders with such a great length, and mostly being situated near the neutral axis, were unimportant as regards longitudinal strength. Ship owners nat- urally objected to these serious objections in the holds, and eventually the Classified Societies, recognizing that these side stringers did not materially support the framing, reduced the dimensions of these stringers, at the same time somewhat in- plate and at the beam knee, which were. considerably improved. Quickly following these developments came the demand for single-deck ships without shelf stringers or hold beams of any kind. While my firm was con- structing one of these ships, materially over 24 ft. in depth, per Lloyds rules, I became greatly impressed with what ap- peared to me a great waste of material in the frame girder, which, in the vessel to which I refer, was 12%4 in. deep by 10/20 to 9/20 in. in thickness by 3% in. flanges and 314 in. overlap. It appeared to me that a considerable reduction could be made in the effectual length of the frame girder, and consequently in the scantlings, and at the same time the dan- ger existing in single-deck steamers of the margin plate cracking would be guard- ed against by working from the tank top as a foundation, and in curving the re- verse frame gradually to a point at a comparatively great distance above the tank top. This suggestion met with such strong approval from the Classifica- tion Societies that it led to the develop- ment of what is known as the "C"' system of framing, on which system my own firm has built six steamers. The frame was reduced to 12 in. by 9/20, with sin- gle attachment angles and no increase in way of long hatches. Owing to the reverse frame being secured so far inboard, there is less liability to trip and there is full justification for the omission of some of the stringers (see Fig. 4). The form of frame naturally BIG. 4. creasing the depth of the frame girder (see Fig. 3). The attention of naval ar- chitects in single-deck ships, and vessels with deep lower holds, was then forcibly directed to the connections at the margin suggested making a similar curved con- nection at the upper end of the frame girder (see Fig. 5), but I have never gone to the extent of curving the reverse frame into the beam knee, as sufficient recogni-