ON THE ELECTRO-MAGNETIC METEOROLOGICAL REGISTER, xli 
The principal regularly recurring actions connected with the clock train 
are two: by means of one the plungers are gradually and regularly raised in 
the tubes of the instruments during five minutes, and are allowed to descend 
during one minute; by means of the other, a small type wheel, having at its cir- 
cumference fifteen figures, is caused to advance a step every two seconds, while 
another type wheel, having twelve spokes but only ten figures, is caused to ad- 
vance one step when the former completes a revolution. The complete revo- 
lution of the second type wheel is effected in six minutes, that is, in the same 
time occupied by the ascent and descent of the plungers. Thus every suc- 
cessive division of the range of an instrument corresponds with a different 
number presented by the two type wheels, the same division always corre- 
sponding with the same number. The two blanks of the second type wheel 
are presented during the return of the plungers, which occupies a minute, 
and during which time no observation is recorded. 
The breaking of the contact between the plunger and the mercurial column 
in an instrument obviously takes place at a different position of the type 
wheels, according as the mercury is at a different elevation ; if, therefore, the 
types be caused to make an impression at this moment, the degree of eleva- 
tion of the mercury will be recorded. It will be seen that observations in 
different half-hours are not made at exactly corresponding instants; but this 
is of no consequence, as the instruments will not sensibly vary within five 
minutes, the greatest possible extent of the deviation. 
I will now proceed to describe the means I employ for recording the num- 
ber corresponding to the degree of elevation of the mercury. To simplify 
the explanation, I will at first suppose the indications of a single instrument 
only are to be registered. One end of a conducting wire is connected with 
the mercury in the tube of the instrument, and the other end with the brass 
frame of the clock, which is in metallic communication with the plunger. In 
the course of this circuit an electro-magnet, such as I employ in my electro- 
magnetic telegraph, and a single very small voltaic element are interposed. 
The electro-magnet is so placed as to act upon a small armature of soft iron 
connected with the detent of the second movement. All the time that the 
plunger is in the mercury the armature remains attracted, but at the moment 
the plunger leaves the mercury the attraction ceases, and the release of the 
detent causes a hammer to strike the types and impress them by means of 
black copying paper on the cylinder. The armature subsequently remains 
unattracted until the plunger descends; immediately before it reascends, a 
piece of mechanism, connected with the clock movement, brings the armature 
into contact with the magnet, which remains there, in consequence of the 
_ recompletion of the circuit, until the contact is again broken. 
It might be thought that a separate striking movement and a separate pair 
of type wheels would be required for each different instrument; but a very 
simple contrivance enables me to register the indications of all the instru- 
_ ments, employing for each by means of the same apparatus. For this pur- 
_ pose a rheotome is so placed in the voltaic circuit as to divert the current 
_ each successive six minutes, so that the circuit shall be completed by a 
_ different instrument. Thus, the barometer is registered during the first six 
_ minutes of the half-hour, the thermometer during the second six minutes, and 
the psychrometer during the third six minutes. Two six-minute spaces are 
left for any other two instruments which it may be hereafter desirable to add. 
It is not necessary that the completion of the circuit should be effected by 
mercury, and there are very few meteorological instruments which cannot 
be applied by suitable modifications to this register. 
It may be necessary to mention another important point in the construction 
