46 
SCIENCE. 
It is easy to understand tnat, with this arrangement, as 
the resistance of the arc becomes greater, the strength of 
the electro-magnet through which the current passes de- 
creases, whilst the other electro-magnet becomes more 
powerful ; so that the arm no longer remains in its vertical 
position, and by its inclination influences the mechanism 
which brings the carbons closer together. The contrary 
effect is of course produced when the inverse action occurs. 
An arrangement which, like the one in question, maintains 
constant the resistance of the arc evidently does not exert 
any effect upon the general intensity of the current in the 
circuit; for the variations of this intensity, outside of the 
lamp, exert always the same effect upon the two electro- 
magnets which control the latter.* The experiment with 
the Foucault lamp gave such good results that I decided to 
undertake the construction of a new lamp as free as possi- 
ble from the defects inherent to the Foucault regulator, 
and to those based upon the same principle. 
Fig. 1. 
After some carefully made experiments with lamps hav- 
ing automatic regulation, and others regulated by hand, I 
ascertained that the latter would give a constant light with 
a much less number of battery cells than was requisite 
with the former. Thus with hand regulators I could ob- 
tain satisfactory results with 24 or even 20 Bunsen ele- 
ments, whilst with the Foucault and Serrin regulators it 
was necessary to employ at least 40. The cause of this in 
the first place is that, in these regulating lamps the move- 
ments communicated to the carbons are always too sudden 
( prompte ) for comparatively weak currents, and, in the 
second place, that these movements are constant instead 
of being proportionate to the intensity of the current pass- 
ing by the voltaic arc. From this I naturally came to the 
conclusion that, in order to obtain a lamp suitable for 
practical working, it was necessary to apply the following 
three principles, which I consider as fundamental : 
1. To maintain constant the resistance of the voltaic arc 
we should not employ a constant mechanical force such as 
that of a spring, but a weak derivation from the main cur- 
rent, parallel to the voltaic arc. 
2. To obtain by means of a special derivation from the 
main current the movements augmenting or diminishing 
the distance between the carbons, in order that the 
rapidity of these movements may be proportionate to the 
intensity of the current producing the voltaic arc. 
3. To make arrangements such that this rapidity of the 
movements communicated to the carbons should, at certain 
periods, be proportionate to the variations in their dis- 
tance ; that is to say to arrange the apparatus so that, in the 
case where the carbons have to be moved towards each 
other through an appreciable space, the movement commu- 
nicated to them may be more rapid than when they have to 
be moved through a very short distance. 
In 1871 I had constructed a lamp which fulfilled the two 
first of these principles, and which was brought before the 
Moscow Society of naturalists. In this system I employed 
♦ This is somewhat obscure ; what is meant, perhaps, is that the sum of 
the currents traversing the two magnets is, with the adjustments adopted, 
a constant value. — E d. E. 
as motor a small electro-magnetic machine of Fromcnt, 
worked by a derivation from the principal current passing 
by the carbons ; and above this electro-motor, the axis of 
which was vertical, were placed the two electro-magnets of 
the differential system above referred to. An armature 
common to both and suspended between their poles like a 
pendulum, reacted upon a double system of gearing, the 
axis of which, furnished with two angle-wheels of unequal 
diameter, would present to the electro-motor one or the 
other of these wheels according as one or the other of the 
two electro-magnets was the more energetic. As the wheels 
in question corresponded to two opposite points of the 
driving wheel, the movements produced were in opposite 
directions and could increase or diminish the distance be- 
tween the carbons with a rapidity greater or less according 
to the intensity of the current, since the working of the 
motor was dependent upon this intensity. The drawing of 
this lamp has been in the polytechnic museum of Moscow 
since the commencement of the year 1873. 
At the end of 1873, M. Jablochkoff, who at that period 
had a mechanical workshop at Moscow, being convinced of 
the superiority of the systems of constructing lamps on the 
derived current principle, made in his workshop a lamp on 
this principle. I shall not refer to the experiments with 
this lamp, which gave full satisfaction to several persons. 
For my own part, I was but partially satisfied, on account 
of its complication, and because it did not fulfil the third 
of the fundamental principles I had laid down. 
In 1874, I arranged a new lamp, the design of which I 
brought before the physical section of the Moscow Society 
of Naturalists, and which is represented by Fig. 1. 
E E' are electro-magnets disposed like those on the other 
systems and having poles, a b, spread out in circular form 
as in the Gramme machine. K is a Gramme or Siemens 
ring, the rotary motion of which causes the carbons to move 
through the intermediary of a double-thread screw, A, and 
two nuts, B C, which carry the carbons. Lastly, D is a 
regulating screw, for the purpose of raising or lowering the 
luminous focus. 
The current passes from the positive pole of the genera- 
tor to the negative pole by three derivations, one of which 
includes the arc and traverses the ring by means of the 
contact-pieces in n ; whilst a second, also including the arc, 
excites the electro-magnet E (or both electro-magnets in a 
given direction) ; and a third which, without passing by the 
arc, influences the high resistance magnet E' (or both mag- 
nets in contrary directions), so that the action of this magnet 
upon the ring shall be in a reverse direction to that of E. 
In consequence of this arrangement the action of the 
electro-magnets upon the ring K is almost nil when the arc 
possesses its normal resistance ; but when the resistance of 
the arc augments the action of the electro-magnet, E be- 
comes weakened, allowing E' to preponderate, and the ring 
K will rotate so as to bring the carbons into closer prox- 
imity. The contrary effect will, of course, be produced if 
the resistance of the arc should diminish. 
