528 A UNIVERSAL METEOROGRAPH. 



of making this transformation appears to me to be what I am about to 

 describe. It should not be forgotten that the construction of my meteor- 

 ograph rests on the hypothesis that the indications of each meteorologi- 

 cal instrument be registered once an hour. The instruments for recording 

 the velocity of the wind and the quantity of rain should therefore be so 

 arranged that the wheel which makes the records shall never make more 

 than one entire revolution in an hour; otherwise it would be doubtful 

 whether the current of air had traversed, for instance, 2 or 22 kilo- 

 metres during the hour elapsed, or whether the rain-gauge had dis- 

 charged 5 or 105 litres of water. 

 Suppose two wheels, A and B (Fig. Ill), of the same diameter, lying in 



the same plane and placed so 

 \f> that the distance between their 



centers shall be exactly equal to 

 the circumference of each. An 

 endless baud or chain, passing 

 around the wheels, connects 



them together. To this chain 



^ TTT are attached three pins, b b b, at 



Fig. III. L ' 



equal distances and consequently 



the length of the circumference of the wheels apart. If, now, the wheel 

 A is put in motion by one of the instruments, this motion will be trans- 

 mitted to the endless chain and to the wheel B. At every stage of the 

 movement one of the pins will be traversing the scale p p, placed 

 underneath one of the straight portions of the chain; in a complete rev- 

 olution of A the pin will travel the entire length of the scale, and 

 when it leaves the scale near B the following pin will immediately re- 

 place it at A. 



We will now pass to the description of the registering apparatus, and 

 take up, at first, that part which is placed in the detached observatory. 



The lever A (Fig. IV) is put in motion by a barometer, B by a metal- 

 lic thermometer, and C by a hygrometer. Each of these communicates 

 its motions to a special wheel, A' B' C, to which is attached the index 

 of the instrument, as shown in the section Fig IV &. These levers and 

 wheels are not situated in the same plane, but placed one above another, 

 so that the wheels turn on a common axis, but perfectly independently 

 of each other. The extremities of the indexes, however, should move 

 in one and the same plane, very near to, but entirely clear of, the divided 

 arc M M INT M, and for this reason two of them are bent. The indexes 

 themselves or at least their extremities, which terminate in a thin fine 

 tip, are of ebonite — a substance which does not conduct electricity. 

 Finally, each of these indexes, even in its widest excursions, moves over 

 but a quarter of the circle, so as not to come into contact with the others. 

 Three of the quadrants being allotted to the barometer, the thermometer, 

 and the hygrometer, I use the fourth for the indications — transformed, 

 as before mentioned — of the wind vane X, the anemometer Y, and the 



