the South African Inatilulion. 109 



This velocity could be attained only by giving the propellers 

 'ery long and cumbrous arms, or else very great rapidity at 

 heir axles. The construction is evidently therefore not suitable 

 to the purposes of ordinary navigation. But as the apparatus 

 may be put entirely under water, it may be found efficient 

 enough, to give a secure but slow movement, to a steam-ship 

 armed to defend a harbour. 



3. The third variety is when b,oth the movement and the 

 impact are direct, as is the case with the piston of a pump, or 

 a vertical surface drawn or driven in the direction of a vessel's 

 progress. A contrivance of this kind is proposed by Captain 

 l'ole, in his second memoir. It may be remarked in general, 

 that when a flat surface is propelled against the water outside 

 of a vessel, the friction of the spindles or rods which carry the 

 surfaces or flaps, and the resistance to the return of these 

 surfaces, when their effective stroke is made, must be so great 

 as to render such constructions quite inefficient for any but 

 very slow movements. A modification of this variety of pro- 

 peller was among the earliest contrivances proposed lor 

 moving ships. It was suggested by BernonetU, and afterwards 

 by Dr. Franklin. It consists of a pump inside the vessel, 

 drawing water in at the bow, and ejecting it at the rf stern. 

 Though never yet successfully put in practice, it seems to be 

 the most promising of those methods which admit of the 

 machinery being entirely under water. The old engine of 

 Savary, without a piston, or with a quantity of oil in place of 

 one, might be employed ; or the piston of the pump might bo 

 attached to that of the steam cylinder; in either case we should 

 have a machine of the simplest and most convenient form con- 

 ceivable, especially if in the latter instance it were so disposed 

 as to work horizontally. A patent for a contrivance on this 

 principle was taken out in 1820. The details in the Edinburgh 

 Philosophical Journal, vol. v. seem to show a misconception of 

 the scale on which au effective apparatus of the kind must be 

 constructed. This principle possess the advantage of allowing 

 the velocity of the striking or pressing surface, to bear any 

 ratio whatever to that of the stream it generates The velocity 

 of the steam piston ought not to be so much as two miles per 

 hour, and if, for the purpose of attaining the effect of 1600 

 mentioned above, we retain the same section and velocity in the 

 stream, each stroke of the pump would need to expell more 

 than would fill five limes the length of the stroke of the steam 

 piston, in a channel having a section of 16 square feet. This 

 could be effected by means of boxes .inclosing a series of 

 pistons attached to the same piston rod, but only by the saciifice 

 of a very great space inside the vessel. If, however, we re-. 



