The reaction with litmus was almost uniformly acid, although seldom strong enough to admit of determination. In a few cases no action was noticed and in a very few a slightly alkaline reaction was observed. It is possible that these alkaline cases may have been due to decomposed urine, all the samples being old (i. e., more than one day).

The quantity excreted varied considerably, but the quantity as given in the table does not show the actual quantity voided. It simply shows the amount available for analysis, and in many instances this was only a small percentage of the entire amount, but from physical reasons it was impossible to obtain any more in a clean condition, free from extraneous substances. The quantity column of Case I fairly represents the actual amounts per day.

The amount of moisture varied considerably and its determination was generally of no use except as a means of reducing the subsequent results to a dry basis. Many of the samples were fairly fresh and in such cases a percentage of about seventy was attained. Departures far from this amount occurred only in old and partially dried samples.

The quantity of fat, casein, and cholesterol varied considerably. Not only is a large variation noticeable between the different individuals but also in the same case from day to day. There seems to be no relation existing between the proportions of these three, except that generally a large amount of fat is accompanied by a correspondingly large amount of casein. This could be expected, as the presence of an increase in one would be caused by a derangement of the intestinal canal and should have an effect on the other also. This is quite evident in In this instance the

the latter part of the data from Case I. removal of the child to the country had a marked and unmistakable effect on the feces; still, the child was reported in good health.

Incidentally, it may be mentioned that the characteristic odor of adult feces was entirely wanting. It showed itself in one instance only, and this case was dismissed for that reason. Subsequent investigation showed that the child had been fed on a mixed diet. Indol and skatol were looked for but not formed except in the case just mentioned.

All the children were fed on the same kind of milk, and given approximately the same quantities per day. This milk was not a single cow's milk but a milk prepared by the "Gaertner Mother Milk" method from the average milk of a large number of cows, and care was taken to have the milk as near as possible of the same composition. The analysis of the milk was approximately

The milk was sterilized perfectly and did not become sour after standing exposed for three days.

Each child consumed from one to one and one-third liters daily and hence ingested

Cases I and II were well cared for throughout the investigation and the results obtained may be considered as reliable as possible from such cases. The others are given to show the range of results obtained, but from reasons beyond the control of the investigation, modifying influences were present; and, while the analytical results are individually correct, the connection between the food ingested and the excreta cannot be traced.

ELECTRIC FURNACES FOR THE 110-VOLT CIRCUIT.'

BY NEVIL MONROE HOPKINS.

Received August 29, 1898.

T occurred to the writer in wiring a couple of experimental arc lamps across the feeders of an incandescent lighting system, that a laboratory electric furnace could be operated on a series carbon plan, without disturbing the protecting fuses of the circuit. The idea of focusing a pair of arcs within a small crucible, or furnace, using only the amount of resistance located in the tops of typical series lamps, proved, however, to be unsatis

1 Read at the Boston meeting of the American Chemical Society, August, 1898.

factory, the current absorbed becoming abnormal, upon introducing a charge for fusion, when its character embodied fair electrical conductivity. To obviate this difficulty, as well as to compensate for the lowering of resistance due to eddy currents between the poles when run in combination with a charge rich in carbon, or graphite, as in the making of calcium carbide, a third pair of electrodes was placed in series, together with a variable rheostat in place of the resistance wire of the lamps. With this arrangement, calcium carbide was readily prepared on a laboratory scale, using only a twelve ampère fuse wire in each leg of the feeding conductors, and allowing the separation of the electrodes through a distance of three and one-half inches. The charge of lime and coke was finely ground together in an iron mill before feeding into the furnace, insuring a homogeneous body. It may be of interest to state in connection with this work that lime and charcoal are poorly adapted for the purpose, the charcoal, because of its floury nature, oxidizing rapidly away from the lime without combining with it. Fig. 1 illus

FIG. 1.-Elevation of "Series Carbon" Laboratory Electric Furnace.

trates the "series carbon" furnace in elevation and partial section, showing a simple construction. Fig. 2 gives a plan view

FIG. 2.-Plan View of Furnace with Cover Removed.

of the apparatus with the cover of the furnace proper removed. This furnace, which consists of an iron shell, lined with fire-clay, should not be over twelve inches long if intended

for use with ordinary electric light carbons, because of their limited reach. The screw mechanism of this "plant" is readily put together by cutting into sections a large furniture-maker's clamp, using the unthreaded portions for the end bearings, and the two threaded sections for the center pieces, which travel back and forth. The carbon electrodes are conveniently mounted on a block by means of tightly-fitting brass tubes, attached in turn with staples or "straps" and screws. Contact is made with the ends of the carbons by pushing brass spring jacks under their ends, to which the connections are soldered. These pieces of spring brass allow of the rapid adjustment of the carbon pencils, or their removal when too short for further use. To start the furnace, when properly connected, it is necessary to feed the electrodes into the furnace until they are all in good contact, and strike the wooden incline planes with a mallet, which causes the ends of the carbons to vibrate or rub together, and make good contact. When the arc has once formed the furnace may be handled like an ordinary single carbon equipment. The method of wiring is given in Fig. 3, where M represents the

A

R

S

12 12

M

FIG. 3.-Diagram of Connections with 55- or 110-Volt Lighting Circuit, with Ammeter and Voltmeter Arranged for Studying the Behavior of Bodies within the Arc.

main feeders, with a fuse block connected at the right, carrying a pair of twelve ampère fuses. The knife switch S is an important feature, and should be within easy reach of the furnace. The ammeter A and the voltmeter V are included as a matter of

interest, and afford a ready means of watching the behavior of different bodies with the arcs. The variable rheostat, R, completes the equipment, which must also be in easy reach of the furnace. In order to avoid annoyance from intermediate fuses, the apparatus should be connected, as near the meter as possible, if the current is sold on this plan. Should one of the twelve ampère fuses melt out, it should be replaced by another of the same capacity, and a little more resistance put in by means of the variable rheostat. Of course, fifteen and twenty ampère fuses may be used if desired, the twelve ampère size being about the smallest for the fusion of compounds possessing carbon as one of the ingredients. Should the variable rheostat at hand prove of too low a resistance for the preservation of the twelve ampère fuses, a second rheostat may be placed in series with it, although with the triple arc arrangement this will seldom be necessary. Fig. 4 illustrates a method for using a sin

FIG. 4.-Method of Using Single Carbons for making Alloys and Small Fusions.

55- or

gle pair of carbons for making alloys and other small fusions. Here, at least, two large rheostats will be necessary, and fuses of larger capacity must be employed. The small crucible is drilled through the bottom, receiving the lower electrode with a tight fit. The outer casing of the furnace in this instance consists simply of a large flower-pot filled in with some poor conductor of