526 A UNIVERSAL METEOROGRAPH, 



Such was my first idea, and, iu spite of the modifications which I have 

 successively made, it still forms the basis of the mechanism which I am 

 about to describe, and which seems to solve, in a simple way, the prob- 

 lem of a complete meteorograph for detached observatories. 



In this description, as well as in the figures designed to make it more 

 readily understood, I neglect entirely the meteorological instruments 

 themselves, because, as I have already said, any instrument may be em- 

 ployed winch is capable of transmitting its movements to a lever. The 

 figures, then, show only the levers terminating in toothed segments 

 which gear into the toothed wheels to which the indexes are attached. 



In all these instruments the motions of the levers should be so regu- 

 lated that the indexes in their widest excursions may only describe a 

 determined arc, the number of degrees in which should depend on the 

 number of instruments desired to be combined in the meteorograph. In 

 the example which I have chosen the indexes need not be separated 

 more than 90° in order to avoid coining in contact with each other, and 

 in order that the indications of any instrument may never encroach on 

 the zone of the cylinder intended for another. The system I have 

 adopted depends, in fact, on the successive observation of the different 

 instruments. 



For records of wind and rain a different arrangement is required. 

 The determination of the direction of the wind requires an entire revo- 

 lution, and so also does the measurement of its velocity by Robinson's 

 anemometer ; in the case of the latter I have provided that the record- 

 ing wheel shall make but one revolution while the current of air passes 

 over a space of 20 kilometres (12 miles), which is seldom less than an 

 hour. 



For the rain-gauge I propose an arrangement which I have recom- 

 mended for many years for the measurement of liquids in general and 

 of granular solids or powders, which figure II will render intelligible 

 without detailed explanation. 



Let A be a reservoir of known sectional area, serving to collect the 

 rain, the quantity of which is to be measured. The water falls through 

 the tube a (at the bottom of which is a watering-pot rose, represented 

 in M, in order that the fall of water may not produce a shock) into a 

 bucket, b b />, which has the form of a sector of a cylinder and can oscillate 

 on the axis <l. As the figure shows, there are two buckets of equal size- 

 coupled together, one of which is always under the rain-tube while the 

 other is discharging the water it has collected. The transverse section 

 of the buckets is a scalene triangle, so that the vertical plane passing 

 through the axis divides the bucket into two parts of unequal capacities. 

 The excess of weight in the outer part tips the bucket, and, in order that 

 the amount of water necessary to produce this motion (for example, a 

 litre or a fraction of a litre) may be regulated with precision, the lips m m 

 of the buckets fall on India rubber cushions o o, the position of which 

 is regulated by the screws E R. The figure shows how the movement 



