Account of some Meteorological Instruments. 315 



their length, which allows their being brought together, just below the prime 

 index, when the instrument is adjusted fora fresh observation. The principle 

 of this modification of the anemoscope is so obvious, and its construction so 

 simple, that we conceive no farther description requisite. 



Should it be considered desirable to have an instrument to indicate more 

 than a single revolution of the vane, a modification of Fig. 2. will produce 

 it ; and Dr Traill exhibited to the Royal Society of Edinburgh, an anemo- 

 scope, which marked four revolutions of the vane. In Fig. 2. the bevelled 

 wheels had each forty-two teeth ; these were retained ; a pinion of twenty- 

 one leaves was fixed on the axis of the vortical bevelled wheel c, to give mo- 

 tion to a small wheel with forty-two teeth, having on its axis another pinion 

 of twenty-one leaves, which moves a second small vertical wheel, also with 

 forty- two teeth; the axis of this last is hollow (to allow the passage of the 

 axis of the vertical bevelled wheel c), and it gives motion to a stud-arm moving 

 round the index-plate, by these combinations, once, while the prime index 

 makes four entire revolutions. This stud-arm moves the indices. 



Dr Traill finds that this apparatus moves pretty freely, and will certainly 

 register all the probable variations of the wind during twenty-four hours. 

 He considers it, however, unnecessary to provide for contingencies so rare, 

 and believes that every useful purpose may be served by the form of the in- 

 strument in Fig. 2, especially if the meteorologist, during very variable wea- 

 ther, would make two instead of a single daily observation. 



II. Register Barometer. The philosophic world is well acquainted with the 

 ingenious contrivance of Mr Keith for registering the variations of the baro- 

 meter ; but the means employed by Dr Traill have the advantage of sim- 

 plicity, and, while the instrument is kept at rest, of not being liable to de- 

 rangement from slight causes. He employs two barometric tubes, arranged 

 as in Fig. 3, and attached to a board C. A is a common diagonal barometer 

 into which is introduced, before the upper part of the tube is bent, a smooth 

 cylinder of iron-wire, half an inch in length and nearly equal in diameter to 

 the tube. The tube is filled in the usual manner ; and on placing the tube, 

 as in the figure, a slight manipulation disengages the cylinder from the mer- 

 cury, and it remains in the void space of the tube. "\Vhen the barometer 

 rises, the mercurial column pushes the cylinder before it ; and when the 

 barometer sinks, this index is left behind ; and therefore its lower end marks 

 the maximum height of the barometric column. 



B is nearly on the principle of what is termed the rectangular barometer. 

 The cistern is placed above, and the aperture is on the other extremity of 

 the tube ; which in order to give both indices similar resistance to the move- 

 ments of the column, is inclined in a degree equal to the diagonal portion of 

 the tube A. An iron-wire index is slipt into the open end of B, after it has 

 been filled with mercury, and is fixed in the position represented at B in the 

 figure. As the atmosphere becomes lighter, the mercury descends from the 

 cistern, pushes the index before it ; and the minimum of the mercurial column 

 is obtained by the position of the index. 



Having remarked that the entrance of dust was apt to impede the free mo- 



