569 



Tliii'd. It should move slowly, so as to require little 

 wheel- work to bring down the space to the size of a sheet, 

 and not be liable to rack itself to pieces. 



Fourth. Should not require to be tui-ned to face the 

 wind. 



Fifth. All made on the same model should tell the same 

 stoiy, without any trial or adjustment. 



The fom- last decide against vertical windmills, and Dr. 

 Robinson used an excellent form of the horizontal one sug- 

 gested to him by Mr. Edgeworth, who once showed liim the 

 covers of a child's globe attached to a rod, and revolving in the 

 wind by the excess of pressure on the concave and convex 

 sm-faces. He suggested that such a miU might be of economic 

 use in many cases, particulai'ly in drainage, as it required no 

 care ; and he referred to experiments by Dr. Corrigan and Mr. 

 Bergin, where it did good work in pumping. He showed a 

 model of it, and explained its action. Hydrodynamics are 

 rather faithless as to impulses on oblique or curved surfaces ; 

 but in this case there appears to be a compensation in the 

 eiTors of the theory, and it gives results surprisingly near 

 those of actual experiment. He gave a sketch of the theory, 

 showed how its constants were determined by experiment, of 

 which in particular the resistance on the concave is four times 

 that on the convex, and stated the experiments wliich gave 

 the relations between the velocities of the wind and the centre 

 of the hemispheres which act as vanes. Both concur in estab- 

 lishing the striking fact, that (except so far as friction inter- 

 feres) the wind moves exactly three times as fast as the vanes. 



He then described the instrument. The four hemi- 

 spheres are a foot in diameter ; their centres describe twelve 

 feet in each revolution. It was necessarily of great height, 

 sixteen feet, to clear the domes of the observatory, which has 

 much increased its weight, twenty-four pounds, and with it 

 the wear and friction. It was an old saying, "as swift as the 

 wind," which, however, will not hold now ! A train only as 

 2 T 2 



