ceiver which contains the rods, interferes with the ease of the experiment in void spaces. It is necessary, therefore, to be content with the length of arc which can be obtained in common air, and to seek only to regulate as far as possible the consumption of carbon, and the distance at which the extremities of the rods must be held in order that the luminous arc produced shall have its maximum intensity. The first contrivances for attaining this end were not successful. The carbon rods were pushed together by the hand in proportion as they diminished in length; but they received not the necessary regularity or proportional quantity of motion. Petrie in England and Foucault in France had simultaneously the idea of applying the electrical current itself to regulate the advance of the carbon points, which were to conduct this same current under the form of a luminous jet. The regulator of the electrical light, by Duboscq, rests upon the same principle as the apparatus of the two physicists just named. In the

regulator of Duboscq, represented in the wood-cut, an electro-magnet, excited by the action of the electrical current which circulates in the copper thread q of the coil B, inside of which is enclosed an iron cone, F, placed in the base of the instrument, attracts into contact a piece of soft iron, K. To this is attached a bent lever, L, which turns at x upon a horizontal axis, and is pressed up by a spring, s, and rests at o against a short lever, having its axis of rotation horizontal. This small lever carries at d a steel nib, the object of which is to check the toothed wheel r. This wheel has a fly, and an endless screw, V, to which a movement can be given by a second wheel, r', the pinion of which is in connection with the great toothed wheel p. The latter contains the mainspring for moving the machine. This great toothed wheel has two grooves of different diameters, the use of which will soon be indicated, and upon which run the two chains h, h', which, after having passed upon the pulleys p p, p' p', are

attached, at E and E', to two copper tubes, in the prolongation of which, at C and C', the carbon conductors are fixed. A small lever, movable by hand, carries a second nib similar to that of the small lever, and serves to stop at will the motion of the apparatus. This whole assemblage of parts is shut up in a metallic cover, the upper part of which can be raised in order to display the pieces of the interior mechanism. The following is the march of the current in the regulator when the carbon points touch, or when the luminous arc is not interrupted.

"For this purpose, assume, as most physicists do, that the electrical current advances from the positive to the negative pole, and suppose that the positive pole of Bunsen's carbon battery is in communication with the clamp R, and the negative pole with the clamp R'. The current, entering at R, descends through the wire q, which an ivory ring insulates at I and i from the brass plate W W', and from the plate and metallic columns which support the mechanical parts of the apparatus. Running the whole length of the wire q of the coil, the current passes into the iron plate F, which constitutes the pole of the electro-magnet. The tube T, which carries the carbon rod C, touches constantly this plate F, and continues the conduction of the current, which, arriving at the place where the carbon points touch, passes from one to the other, traverses the carbon rod C', ascends its tube T', descends through the column S, and goes from this column to the clamp R', which corresponds to the negative pole of the battery. The column S is insulated from the rest of the apparatus by an ivory ring I', in order that the current, to complete its circuit, may be forced to go through the carbon rods. Things being thus disposed, the carbon points touching, or being removed by the distance best adapted to the most brilliant arch of light, the plate F will be strongly magnetized by the action of the current; the iron K will be drawn into contact; its levers will rest against the toothed wheel r, and, even when the spring in the wheel P is wound up, all will remain in equilibrium in the interior of the apparatus. The electricity will continue to pass, wasting the carbon C by the transfer of its molecules to the carbon C', and both of them by their lively and rapid combustion in the air. After a certain time, the carbon points will be so much separated that the current experiences a considerable resistance in breaking through the intervening space. The intensity of the current being diminished, the spring s will destroy the contact of K; its levers will discharge their function; the tooth d will quit the wheel r; the main-spring will

set in motion the wheel P, and the regulating parts r, r', which are attached; the chain h', which moves the carbon-holder of the positive pole, rolling up on the groove of p', will make the carbon C ascend, while the chain h, unrolling from about the groove of p, will make the corresponding carbon C' descend. The ratio of the diameter of the pulleys p, p' can be changed by a special system of elastic pressures belonging to the groove p of the pulley of the negative pole, and which is represented on a large scale in Figure 2. By the help of a key, any dimension wished is given to the pulley p. This dimension must always be such that the point of contact of the two carbons shall be maintained at the same elevation, in spite of the more rapid waste of the positive pole. Now the quantity by which the negative carbon descends, and that by which the positive carbon mounts, are proportioned to the circumferences of the respective pulleys, and these circumferences are in the same ratio as their diameters. If the positive carbon is wasted three times more rapidly than the negative carbon, the pulley p' must have three times as large a diameter as the pulley p, in order to maintain the point of junction at a constant level. The ratio of the diameter of the two pulleys must be regulated by trial every time the carbons are changed, since a difference in their diameters or their densities may cause a great difference in the waste of the two incandescent extremities. The contact-maker K is provided with a driving-screw, so as to alter at will its distance from the electromagnet, according to the energy of the pile used, which struggles with more or less force against the spring s, which resists contact. The place at which it is best to stop the contact-maker K is easily discovered by a hissing which is produced when the carbons are too near. This screw is so turned as first to provoke this hissing, and then turned gently in the opposite direction until the noise ceases. The proper place for experiment is where the hissing stops. If within or without this position, the carbons are too near or too distant.

"The tube of the negative carbon is provided at n with a nut or an articulated knee, by which there can be impressed upon it, with the aid of the buttoned stem m, a slight conical movement around the vertical, so as always to make its point coincide exactly with that of the positive carbon. After one or two experiments with this apparatus, the play of its parts presents no longer any serious difficulty, and its manipulation becomes as simple as that of the Carcel lamp, or the ordinary lampe à moderateur."

Professor Lovering asked the attention of the Academy to a discussion which seems to have been going on as early as 1843, and which had been recalled to his notice by a letter which he had recently received from a friend in Cincinnati, who questioned the propriety of including among the questions for the examination of candidates for the High School of that city the following: - "Does the Mississippi River run up hill or down hill?"

"I shall introduce what I have to say upon the subject with an extract from the Common School Journal.* The article from which I make the extract is headed Geographical Error.' The writer

"The following egregious blunder, with the captivating title Water running up Hill,' is going the round of the public papers, to be caught up by thousands of school-teachers, and imprinted upon the minds of tens of thousands of scholars. Dr. Smith, in a recent lecture on Geology, in New York, mentioned a curious circumstance connected with the Mississippi River. It runs from north to south, and its mouth is actually four miles higher than its source: a result due to the centrifugal motion of the earth. Thirteen miles is the difference between the equatorial and polar radius; and the river, in two thousand miles, has to rise one third of this distance, it being the height of the equator above the pole. If this centrifugal force were not continued, the rivers would flow back, and the ocean would overflow the land.'

"This statement of Dr. Smith, when separated from the paradoxical declaration with which the newspapers have heralded it, is wholly correct, except in the numerical details, in which Dr. Smith evidently did not aim at great precision. But the writer in the Journal (who is understood to be President Horace Mann) not only attacks the accuracy of these details, but assails the mechanical principle which lies at the foundation of Dr. Smith's statement; saying, that it would be difficult to compact a greater number of errors of fact and of principle into one short paragraph, than are found in the above quotation.' The precise numbers involved in this question are of secondary importance. I am willing, and Dr. Smith no doubt is willing, that Mr. Mann

* Vol. V. p. 65.

should have the numbers as he states them. Suppose then that the length of the Mississippi River, measured on a meridian, is only fourteen hundred miles, and that the mouth is only about two and a half miles more distant from the earth's centre than the source. The question arises whether the flow of the river from north to south is caused by the centrifugal force, or whether the criticism of Mr. Mann upon this mechanical solution of the problem is sound. The critic asks: Why then does not the mighty force which sends the Mississippi up hill four miles send the Nile back to the Mountains of the Moon?' And again he asks: Why does not the centrifugal motion of the earth drive the waters of the Atlantic and Pacific Oceans towards the equator, at the rate of ninety-six miles a day?'

"Let us attend next to Mr. Mann's own explanation of the flow of the Mississippi. After enlarging upon the protuberant matter at the earth's equator, he continues: Now water, like every other material thing, is attracted towards the centre of gravity. The centre of gravity is that point about which all the parts are in equilibrio. Or, in popular language, water, like everything else, being attracted by matter, is most attracted, other things being equal, by the greatest quantity. The only philosophical idea we can have of up or down is from or towards the point of greatest attraction, that is, from or towards the centre of gravity.' Elsewhere, this writer speaks of the earth 'being an oblate spheroid, having the greatest quantity of matter, and therefore the greatest attraction, under the equator.' Finally he says: The whole truth is, that the waters of the Mississippi are constantly tending to the common centre of attraction; but, being prevented from approaching that centre in a direct line, they approach it indirectly, by moving forwards along the bed of the channel. They are constantly approaching the centre of gravity, that is, they are constantly descending.'.

"One error into which Mr. Mann has fallen is that of supposing that the attraction which the earth exerts at any particular point of its surface is a local phenomenon, and not the resultant of the aggregate attractions of every particle of matter in the earth. This error leads him to a conclusion contradicted by the experiments and observations of the last two centuries; namely, that where there is the most matter, there is also the most attraction, and that consequently the attraction is stronger at the equator than it is at the poles. We might ask Mr. Mann why this mighty force of attraction does not send the