178 



NATURE 



[December 21, 1893 



do not vary more than eight seconds from the mean out 

 of a total of 176 seconds. This remarkable precision has 

 been always found. It was equally present in a steam 



experiment, would make each ship turn towards the 

 other, X turning to the right, and Y turning to the left; in 

 which case it would not be possible for the ships to touch 



No. I. — 17° Helm."! 55 Revs, 80 to 91 lets. 



No. 2. — 24^° Helm.! Dft. of water. 25"ioF. 27*4 A. 



No. 3. — 33J' Helm./ May 12, 18" 



Fig. 10. — To show effect of Helm-angle. Scale, jl inch = 100 feet. 



pinnace, and the one experiment made with a very light 

 ship in a high wind failed to disclose any difference due 

 to differing directions of the wind. ' 



The traces of the Edinburgh and Dreadnought are 

 brought together in order to exhibit the wide differences 

 that exist in the form of the path described by different 

 ships in turning. They show how imperative it is that in 

 fleets, at any rate, these differences of form should be 

 recognised. 



But the peculiarity of the form of the path remainsnearly 

 the same at all helm-angles, and this makes the necessary 

 equalising of the paths for the purposes of fleet manoeuvr- 

 ing easier. It shows, too, the fallacy of the 1865 idea — 

 still preserved — ^tliat there can be a single '' evolutionary 

 helm-angle " suitable to equalise a large or a small turn. 

 The facts are illustrated by Fig. 10, which shows the 

 effect of varied helm on H.M.S. Thunderer. 



I have now only to point out how experiment bears on 

 the question of collision between ordinary ships at sea. 

 The ordinary form of approach before collision is given 

 in Fig. II. The law enjoins that Y should keep steadily 

 to her path, and that x should " keep out of her way." 

 In order to do so, she has been for something like thirty 

 years told that she must decide for herself whether to 

 turn to the left or to the right. Needless to say that as 

 she cannot hope to turn " sharper" than the path marked 

 for her, she generally produces collision by turning to the 

 left, but she is never explicitly condemned for that act. 



Ships often only discover one another when so close 

 that it cannot be certain which has the power of avoid- 

 ing collision. Fig. 12 supposes two Ediitbm-ghs meeting 

 under such conditions, where it is seen that safety can 

 alone lie in knowledge of the general manoeuvring powers 

 of ships and their application to the particular case. 

 Admiral Beechey's law would have caused both these 

 ships turn to the right, and would have made collision 

 inevitable. The existing law — if it were acted on— 

 would compel Y to keep her course, in which case also 

 collision would be inevitable. The natural law, based on 



1 I have by me many scores of experiments made with fifteen or 

 .^i-xteen ships and vessels of all sizes and classes. 



Fig. II. — Scale, (say) i inch = 100 yards. 



NO. 1260, VOL. 49] 



The figures denote ^r each ship, eoual 

 intervals of timt. whether iurninq ornot 

 At fZ knots initial speed the interval is 

 <i.ve seconeis en this scale 



Fig. 12.— Scale, '288 inches = 



100 yards. 



