EXPEIUMENTS ON SURFACE-FRICTION. 123 



It was also necessary to eliminate certain other constant resistances known 

 to exist, namely, that due to the air-resistance on the swinging bar, that duo 

 to the excess of surface of the cutwater, owing to its projecting up through 

 the water above the upper edge of the planes, and that due to the projections 

 and irregularities on its surface, caused by the fastenings of the planes. Of 

 tliese, the air-resistance was obtained by direct experiment; that due to 

 excess of surface was calcidable on the data already possessed ; and the 

 resistance due to the projections <fec. was determined by trying the 1-foot 

 G-inch plane, with its surface smoothed up Avith paraffine and varnished as 

 before. The deduction of these constants brought down the line to the plain 

 lines shown on Plate VII.* 



But the first-mentioned object, that of deciding the friction of very short 

 lengths, I have so far been unable to treat quite satisfactorily, owing to the 

 difficulty of guiding very thin blades. I have, however, obtained good 

 results with a 12-inch blade and a 6-iucli blade (see Plate III. d, e) sharp on 

 both edges, both about similar in longitudinal section to the 1-foot 6-inch 

 plane ; and the experiments with these gave spots through which the curves 

 on Plate VII. were drawn for the first 1-foot 6-inch length of surface. 



And though, in the absence of any successful experiment with blades of 

 different thicknesses but the same length, we can scarcely regard as disproved 

 the existence of possible body-resistance due to the thickness, slight as it was, 

 of the planes tried, it is obvious that it would be difficult to deduct further 

 from the diagram of resistance any considerable constant represer iing this, 

 without making the friction per unit of surface decrease with increasing 

 length less in the first 6 inches than it would be uacurally expected to do ; 

 in other words, without making the curvature of the lines on Plate VII. less 

 sharp at their origin than would be expected, seeing that in the rest of the 

 diagram the curvature becomes rapidly flatter as the lengths of plane become 

 greater ; but indeed the thinness of the planes and the smallness of the : e- 

 duction of resistance which followed the substitution of knife-like for rounded 

 edges render it almost impossible to credit body-resistance with any appre- 

 ciable item in the account. It is also most desirable to extend these experi- 

 ments to greater lengths of surface than I have been able to try with this 

 apparatus. But it would indeed be almost impossible to do so in the experi- 

 ment tank ; and I shall endeavour to organize some arrangement by which 

 greater lengths may be successfully tried in open water. 



I have thus far confined myself, in the description of the result, to the 

 question of the eff'ect of lengths of surface upon resistance. I have now to 

 deal with the question of quality of surface. 



The diiferent surfaces tested may be enumerated as follows :— 



Shellac varnish. 



Hay's composition. 



Peacock's composition. 



Tallow. 



Glue. 



A smooth metal surface obtained by a coating of tinfoil. 



The comparison between the first three named was made with planes 5 feet, 

 16 feet, and 50 feet long, which were each coated first with Hay's and sub- 

 sequently with Peacock's composition, all the planes having been previously 



* It should be noticed, however, that the scale of resistance shown on Plate VII. gives, 

 not the actual resistances due to the planes tried, but the reduced resistance due to a, 

 surface one foot wide and of the lengths shown, 



k2 



