162 APPENDIX. 



difference which has been made in the dimensions, in order to convert Cave's wheel into 

 Morgan's, is that the excentric collar F has been made much smaller than it was possible 

 according to the other construction, and this diminution tends to equalize the motion of the 

 floats ; but what equalizes it still more is that a driving rod of a constant, and always greater 

 length, has been substituted for one which was always varying. Another difference, which 

 could not appear in the figures, and which, in our opinion, operates in favor of Morgan's 

 wheel, is that the excentric in Cave's is outside, and in Morgan's inside the framing of the 

 wheel. 



Some remarks on Mr. P. W. Barlow's calculations, &c. 



Very little can be said regarding his mode of determining the centre of pressure of floats, 

 since he has assumed, and not calculated its position ; or rather, he has assumed an empirical 

 rule (see page 45), which facilitates the calculation very.much, but clearly cannot suit all cases. 

 In the following page he assumes the centre of pressure in Morgan's wheel to be situated at 

 one-eighth of the depth of the float below its centre. We think he allows too much for the 

 lower part acting during a longer period than the upper ; for until the point which he con- 

 siders as the centre of pressure enters the water, the power expended on the immersed portion 

 has amounted to scarcely any thing. This trifling quantity being neglected, the resistance in 

 the ensuing portion of the stroke is increased materially, being considered as equal to what it 

 would be if the whole surface of the float were acting at the centre of pressure, while at first 

 only three-eighths of it are actually immersed ; and as the difference between the effective 

 velocities of different parts of the float is not great, it appears that much more resistance is 

 added after the centre of pressure has entered the water, than was neglected at the commence- 

 ment of the stroke. We do not, indeed, deem it improbable that the centre of pressure is 

 situated as high, or even higher than the centre of the float. 



The method of finding the whole power exerted by the engines adopted, (page 50), can 

 seldom, if ever, give a correct result ; for the number of floats acting at one time varies very 

 much in different wheels, and in the same wheels at different immersions ; that is, the mean 

 number varies, for there may be in two cases sometimes three, and sometimes four in the 

 water at the same time, and yet the mean number may be different : in one case it may be 3^ 

 and in the other 3| ; and, as Mr. Barlow assumes it either at 3 \ or 4, the error may be -fe or 

 -fij. of the true power exerted, of which the mean number of floats acting at the same time is a 

 factor. Such an error cannot surely be considered of no consequence, particularly as errors of 

 observation are inevitable in such experiments as those made to ascertain the speed of steam 

 vessels. A fact which proves Mr. Barlow's calculations of power to be inaccurate, is that the 

 proportion of the power expended on the vertical paddle in the experiment of the ' Sala- 

 mander,' differs from the calculated proportion by more than 72 per cent. This is the very 

 example we have worked out (page 127), and found the calculated to agree with the nominal 

 power in a remarkable degree, although calculated according to the same theory as that 

 adopted by Mr. Barlow, but without any arbitrary assumptions, which are probably in a 

 great measure the cause of the discrepancies observed by him, and of his being unable to 

 account for the whole power with Morgan's wheel, though this is, perhaps, principally to be 

 attributed to his confounding it with Buchanan's. 



