54 ECONOMICAL CARGO SHIPS—SOME MODEL EXPERIMENTS. 
to be well stocked. I say practically devoid, advisedly, because there always was some 
information available. ‘The field of fine models, high-speed vessels, war ships and pas- 
senger liners has been well explored before cargo models were taken up seriously and 
consistently. We must thank investigators like Dr. Sadler, Admiral Taylor, Mr. Robert- 
son, also Messrs. Froude, Baker and Semple on the other side, for the large amount of 
model-tank information which is now available in the field of slow-speed cargo-ship 
models. 
Cargo boats used to be dismissed with a wave of the hand as being more or less be- 
neath the notice of such austere institutions as experiment model basins, but the com- 
mercial world can congratulate itself that such is no longer the case. Mr. Robertson’s 
present paper is another valuable contribution to our knowledge. 
With reference to Professor Sadler’s remark that the full shape is not necessarily the 
hardest driven, we have a case in point from experience at the New York shipyards some 
five or six years ago; we thought we had a bow that was too full—that is, some of us 
thought it was too full, but others thought it was all right. We tried the full bow and 
got certain results; we then cut it down to meet the ideas of the critics and tried it minus 
200 tons of displacement. We got identically the same results up to some three knots 
faster than the ship was to be engined for; it is scarcely necessary to say that we built to 
the fuller bow and obtained the extra 150 tons or so of deadweight. The only thing the 
tank did not show up was her behavior at sea, as affected by the fuller bow. But we have 
had no complaints, in fact, nothing but praise, as to the seagoing qualities of that ship, 
besides several orders for duplicate vessels. 
Mr. Robertson’s figures throw light on the different methods of calculating results at 
the National Physical Laboratory (England), the Washington tank, and the tank at Ann 
Arbor, Michigan. Those of us who used figures from these several tanks have known 
for some time that the results did not check up well when comparing figures for slow- 
speed ships. This was sometimes very noticeable on checking a set of figures obtained 
from Washington data with the same ship figured by Ann Arbor methods. I think we 
can congratulate ourselves that this source of uncertainty is about to be removed. If 
you have a correcting factor, such as Mr. Robertson gives, which enables you to interpret 
the results worked by the method in vogue at the different tanks, you do not mind where 
you get your figures, because you can bring them to a constant basis and make them 
comparable. 
Professor Sadler asked for some comment on the Froude constant method as com- 
pared with the pounds per ton. I think Mr. Robertson has made the constant method 
considerably more acceptable to those engaged in the practice of naval architecture by 
eliminating what are generally called the circle P values—that is to say, he plots his 
results on a base of speed-length ratio. If you eliminate circle P, you eliminate half of 
the terrors of the constant system. Otherwise I think the direct method of the constant, 
giving the whole resistance at once, is preferable for everyday work. For study of model 
forms, comparing one with the other, it will always be necessary to keep skin and resid- 
uary resistances separate. For routine work the value of the whole resistance at one step 
is acceptable; but there will be many cases, such as passenger liners, also high-speed bay 
and sound boats, where the skin and residuary resistances must be studied separately. 
There is no doubt that this paper adds appreciably to the knowledge now available 
to the designer of full form vessels. 
