dressed that a light blow would show about one-half an inch in diameter; while the cross-face should show a blow threequarters by three-eighths inch. A sharp, cutting blow is not effective in either knocking down a lump or stretching the metal.

It is always advisable for beginners to start in with a small circular cross-cut saw, one that can be easily handled, and to practice on this until expert.

Current for Annunciators

QUESTION: Isn't there some way to transform a 110-volt current to 10 volts for use in connection with annunciator systems, without using a motor-generator or dynamotor? I always believed one could insert resistance in the circuit and get the same results as the drop in a circuit.

one

one

an

We have an electrician here who has offered to make a resistance box and give us 10 volts from 110. Accompanying is a diagram recently under discussion, in which two lights, 16-candle-power, and a 8candle-power, are in parallel on a 110volt circuit and in series with a push button and annunciator bell. What voltage and what amperage do I get at the bells?—W. L. G.

Answer: It is, of course, perfectly possible to cut down the voltage of a circuit any desired amount by the insertion of resistance therein. This, however, is not a process of transformation, in which the current is changed from one voltage to another without loss, but consists in absorbing the difference between

OBCPO16C.P.

DIAGRAM OF ANNUNCIATOR CIRCUIT.

the original voltage and the desired voltage in a system, which of course represents dead loss and is neither efficient nor economical.

It is perfectly possible to cut down a 110-volt current to 10 volts by the insertion of the proper resistance.

In regard to the conditions shown in your diagram, we should say that in order to tell you the current flowing or the voltage across the bell, it would be necessary

Current for Heating-Voltage and
Power Expenditure

QUESTION: If it requires one ampere at 100 volts to heat a piece of wire 10 feet long, does it not follow that one ampere at 1,000 volts would heat a piece of wire of the same diameter 100 feet long to the same temperature? Does it not require more power to produce a current of 1,000 volts than it does a current of 100 volts; and if so, how much more?— H. E. V'.

Answer: If it requires one ampere at 100 volts to heat a piece of wire 10 feet long to a certain temperature, it follows that one ampere at 1,000 volts would heat a piece of wire of the same diameter 100 feet long to the same temperature. In other words, it is the current which determines the heating; and if one ampere flows, the heating will be the same regardless of the voltage; but you must bear in mind that if you make the length ten times as great, then you increase the resistance in like proportion; and the voltage must be similarly increased in order that the current may remain the

to produce a current of one ampere at 1,000 volts, as it does to produce a current of one ampere at 100 volts.

Equalizing Generators

QUESTION: I have four direct-current generators, two 400-K. W. and two 250-K. W. We have a great deal of trouble getting them to equalize. They will not run together. Sometimes one will lose its load, and then another. Please let me know what the trouble is?-N. H. L.

ANSWER: It is very difficult to diagnose a case of this kind, without tests or extended observation of the action of the machines and their prime movers. The best way, perhaps, is to enumerate the reasons causing machines to refuse to divide their loads properly. The operator can then determine for himself which is the one that applies to the case under consideration.

First, the resistance of the leads from the machine brush or attachment point of the equalizer to the bus-bar must be inversely proportional to the output of the machine, so that, with normal fullload current flowing through them, the drop of potential over each will be the same. If they are not so proportioned, the machine whose resistance is too high will refuse to take its proportion of the increase of load, if equalized at low loads, and will refuse to drop it when equalized at high loads. The resistances of the leads from the brushes to the switchboard on the series side of all machines, should be tested by the fall-ofpotential method; and if any one of the resistances is found to be too low, it must be increased. This can be most readily done, perhaps, by inserting in this lead. an iron bar or some other conductor of a higher resistance than copper; or, if practicable, a smaller lead may be used on this machine.

Second, the regulation of the machines may be different and that of the prime movers the same. This should be determined by noting the rise in voltage from no load to full load when each machine is operating alone. If one machine. is compounded more than another, it will take more than its proportion of the increase of load when equalized at low

loads and will drop its load too quickly when equalized at full load. The remedy for this is to lower the resistance across the series terminals of the machine in question. This is only a partial remedy, however, as this also shunts the series coils of all the other machines to a certain extent. The shunting of the series and the adjustment of the lead resistance must be independent of each other.

Third, the equalizer resistance may be too high, or some of the connections may be bad. The lower the equalizer resistance, the greater will be the tendency of the current to seek this path, and the more sensitive will be the division of the load. For this reason the equalizer should be as short as possible and of ample crosssection. If this resistance is too great, first one machine and then another will tend to take the load or throw it off at the slightest provocation.

Fourth, the magnetic densities of the machines may be different; and the one with the higher density will not respond as readily as the others to the influence of the equalizer current, and will refuse to take its proportion of the load increase when equalized at low loads, or to part with it when equalized at high loads. The test for this is to take the saturation curves of the machines, and compare them. If it is found that, when plotted to the same scale, the normal voltage comes higher up on one machine's curve than on the others, its speed must be raised until the curves are approximately the same. If the machine is belt-driven, another pulley can be substituted.

Fifth, the regulation of the prime movers may be different. Load each unit up separately, and adjust the series shunts so that the rise of E. M. F. from no load to full load is the same in each machine. This makes the regulation of the prime movers and generators together, the same, instead of that of the generators alone, as in the second case.

This should eliminate this cause of trouble, if the load is one that does not vary suddenly through wide ranges as does an electric railway load. In this case the various engine governors are

apt to be erratic in their action relative to one another. This is especially true if the engines are of different types. One governor will not respond as quickly to load changes as another. This will manifest itself in a seesawing of the load from one machine to another without any apparent law or reason. There is no positive remedy in this case. A great deal, however, can be accomplished by adjusting the engine governors.

Sixth, the belt on one of the machines may be slipping. In this case, as in the last, the shifting of the load will be very erratic, and in general the machine will refuse to take its load, and no amount of manipulation of its rheostat can make it do so. If this is at all marked, it will be indicated by the heating of the pulley, and, of course, can be tested out with a speed-counter.

The position of the brushes has some slight influence on the load distribution. The machine with the greatest brush lead will have the largest armature reaction, and its E. M. F. will not rise as much as that of the others with increasing load. It will refuse to take its share of the load increase when equalized at low loads, or to part with it when equalized at high loads.

The ammeter may be connected in on the wrong lead, and so will not indicate the true output of its machine. The ammeter, as well as the circuit-breaker, should always be in on the lead of the machine not containing the series winding.

=

=

in which Q Quantity of water discharged, in cubic feet per second; g Gravity (32.16); b Gravity (32.16); b = Breadth of the notch, in feet, commonly called the "length" of the weir; and H = Depth of water on lower edge, in feet. This depth must not be measured in the plane of the weir, but must be taken a considerable distance back of the weir, on account of the decrease in head where the water flows over.

Measuring Flow of Stream Question: What is the method of measuring the flow of a small stream?

The result obtained by the above formula must be multiplied by coefficient of discharge, in order to obtain the actual discharge. Where a dam already ex

Answer: The best method is by the use of a weir constructed of plank and built into a temporary dam of earth. This method can be used for any small streams of about one to four feet in depth, and up to about 50 feet wide-that is, when they are normal. At times of flood, when the flow is much greater, measurements cannot be made in this way. The formula to use is Q = √2g × b H√H,

DIAGRAM ILLUSTRATING METHOD OF MEASURING FLOW OF A SMALL STREAM.

ists in a stream, observations of the flow over such a dam will give fairly good results, using the coefficient of discharge.

Where a weir cannot be used, then the flow must be determined by actually ascertaining the mean velocity of the flow at a given section, and the area of such cross-section. Then the discharge will be equal to the product of these quantities. Owing to the disturbing effect of the bottom and sides of the channel, the velocity of the water will not be the same at all points in a given cross-section. Therefore, instead of trying to get the average velocity through the entire crosssection, it is usual to divide the crosssection of the stream into several vertical strips, by lines 1, 2, 3, etc., as shown in the diagram, and then obtain the average velocity and discharge of each of the strips a,, a, a, separately. In doing this, a place should be selected where the flow is as uniform and the channel as regular as possible. The area of crosssection is obtained by making careful soundings, and then making an accurate plot of the cross-section, adding together the areas of each section as computed. The velocity can be obtained by timing the movement of a float.

Observatory on Mount Wilson

BY

Y the generous gift of the sum of $150,000 from the Carnegie Institution, there is now assured to California one of the greatest benefits provided in recent years for the extension of astronomical research.

This princely gift will be expended within the next year upon a solar observatory to be erected on the summit of Mount Wilson, near Pasadena in Southern California. This mountain towers 6,232 feet above tide-level. The sum will represent only a part of the observatory equipment, which will cost in total about $300,000.

One of the instruments which will be used on Mount Wilson is a reflecting telescope about 200 feet in length, which

was the gift of Miss Helen Snow to the Yerkes Observatory. Other instruments weighing in the whole more than 200 tons, have been brought from Chicago to Mount Wilson, and many of them will be retained there after the unnecessary ones have been returned.

The atmospheric conditions on Mount Wilson are considered unusually good for observations. Much interest is felt in the new project, through which it is believed that many interesting discoveries regarding the sun's heat and composition will be made.

The sun's spots develop from minimum to maximum activity once in about eleven years, and during 1905 were at their highest development. They will probably be in position for the closest and most fruitful observation during the present year. Prof. George E. Hale, Di

In producing wind-made electricity, Wilson calls upon the windmill to perform its customary function of pumping water. He leads the water into a hydraulic regulator built on the principle of a water-lift, in which the pressure is controlled by weights. Approximately, a uniform head pressure of 75 pounds will correspond to the capacity of the water pumped by a 10-foot windmill wheel. This is increased to 100 pounds for a 14foot wheel. The water is discharged from the hydraulic chamber by means of

er who

has a windmill on his grounds can enjoy electric lights and the many other services which electric power is capable of yielding. For many years, men have been trying to convert wind power into electricity. R. W. Wilson of Westfield, Ind., has worked out a practicable method of accomplishing it. The contrivances heretofore tried for this purpose have usually failed because of their inability to control the variability of the wind power. Generally this failure has been due to effort to derive electricity as a direct product of wind power. Under the Wilson method, electricity is generated as a by-product in the course of the windmill's service in

automatic valves. This regulator is the means of maintaining an even pressure under all conditions, whether the windmill is revolving fast or slow.

Under the uniform pressure, the water is passed from the hydraulic chamber through a water motor to which a dynamo is attached. Then it is discharged through troughs, and led away to the fields if desired; or it can be stored up in tanks or reservoirs to be pumped back into the hydraulic regulator again, in case water economy should be necessary. By producing an evenness of pressure in this way, the dynamo is run at a uniform speed, whether the wind is blowing a gale or is just enough to make the wheel