320 



NATURE 



IJuly 31, 1879 



METEOROLOGICAL REGISTERS'- 



THERE is now scarcely a meteorological observatory 

 ■which is not prorided with registering instru- 

 ments. The number of these is already considerable, 



Fig. I. — Registering Earometer. 



and the methods of their construction are almost as 

 numerous. At the observatory of Montsouris I have pre- 

 ferred continuous registration to registra- 

 tion by points, and the tracings effected 

 by a fine metallic point on glazed paper 

 blackened with smoke to the use of the 

 pencil or to the gau/rage by pressure or 

 by shock. 



The barograph of Montsouris is a baro- 

 meter-balance devised by P. Cecci, and 

 latterly adopted by P. Secchi. The 

 barometric tube A b (Fig. i) is of iron, of 

 3 centimetres internal diameter, and of 

 the same calibre throughout its length. 

 It is fixed, and has, at its lower ex- 

 tremity, a cylindrical plunger, the full 

 section of which is equal to the vacuum 

 section of the barometric tube. The 

 cistern, C, is of steel, and suspended to 

 one end of the beam of a balance, d, the 

 other end of which carries a counterpoise. 

 The centre of gravity of the weighted 

 beam coincides, as nearly as possible, 

 with the central knife-edge, so as to render 

 the equilibrium indifferent. Hence it 

 results that when, in consequence of a 

 rise of the barometer, a part of the mer- 

 cury rises from the cistern into the tube, 

 the former, thus rendered lighter, tends 

 to rise, with the effect of introducing 

 into the mercury of the cistern a greater length of 

 the cylindrical plunger. Equilibrium is only re-estab- 

 lished when (the volume of mercury displaced by the 



" Paper by M. Mariif-Davy, ypurrtal dc Physifue, April, 1879. 



plunger being equal to that which corresponds in the 

 tube to the increase in height of the barometric column) 

 the level of mercury in the cistern returns to its constant 

 point, and when, consequently, the apparent weight of 

 that cistern has resumed its fixed value. The needle, h, of 

 the beam of the balance marks its displacements on the 

 surface of the blackened cylinder, K. It only remains to 

 read the indications. To do this as accurately as possible, 

 the support of the cylinder carries an electro-magnet, the 

 armature of which is provided with an arm bearing a me- 

 tallic point.^ In ordinary weather this point is fixed, and 

 traces on the cylinder a datum-line from which the 

 ordinates of the barometric curve are measured ; but 

 every hour the electro-magnet is acted on by a clock 

 marking the" time on all the registers; the point departs 

 momentarily from the line, and makes a clear stroke, 

 which is reproduced at the same instant on the other 

 cylinders. At the end of each week the cylinder is taken 

 away, to be replaced by another kept ready, and is placed 

 on the reading apparatus. 



This apparatus (Fig. 2) consists of a horizontal steel 

 bench,/_^, provided on its upper surface with two columns, 

 kh', intended to support the axis of the cylinder, and on 

 a vertical stand with two microscopes, which may be 

 moved either separately or together in a horizontal direc- 

 tion. By means of an endless screw a movement of 

 rotation on its axis may be communicated to the cyhnder 

 for the purpose of presenting successively to the micro- 

 scopes the various faces of the cylinder. The micro- 

 scope b is always pointed on the datum-line, and it is 

 to follow this line in its accidental displacements, that 

 the two microscopes may be moved together by the action 

 of an adjusting screw. The microscope which points to 

 the traced line carries at its focus two crossed threads, 

 one horizontal and the other vertical. At each hourly 

 reading one of the horizontal tracings on the datum-line 

 is brought under the horizontal thread of the microscope. 

 But, as the barometric curve often presents inflexions of 

 which it is useful to note the exact time, the critical point 

 of this curve being brought under the movable micro- 

 scope a, a second horizontal thread, movable by a micro- 

 metric screw, serves to measure the distance of this point 

 from the hourly mark immediately preceding. 



Fig. 2. — Micrometer or re.iding curves traced by^ the register. 



The microscope a is supported at the extremity of an 

 arm of a lever the length of which is equal to that of the 



' This electro-magnet, which is fseen in [the thermcgraph (Fg. 3), is not 

 shown in Fig. i. 



