Observations of these stars are available for 130 years. From a discussion of these, the conclusions of Mr. Stone are: 1. "The four stars of the group have proper motions much larger than the average." 2. "They have a common proper motion of more than a second of arc." 3. "Each star of the group is moving away from every other star of the group, by quantities which are small compared with the common proper motion of the group." 4. "That, roughly speaking, the velocities of separation are larger, the larger the present angular separation of the stars." It appears to Mr. Stone very probable that these stars at a remote period (more than 300,000 years, for example) really formed a system or group. Such a group might have been formed by the close approach of two binary stars. Though the conclusions are necessarily doubtful, the case is interesting.New Stars. In November, 1876, Dr. Schmidt, director of the Athens observatory, discovered a new star of the third magnitude in the constellation Cygnus. Its brilliancy diminished rapidly at first to below visibility to the naked eye. In 1877 it had further diminished to 10 magnitude. The most interesting results have been derived from spectroscopic observations of this star. Its spectrum was at first twofold. On a continuous spectrum, analogous to that of the sun, and no doubt due to the light of the solid or liquid photosphere, there was superposed a second spectrum of bright lines, which was due to light coming from incandescent gases, notably hydrogen. The hydrogen lines were at first very brilliant; with the decrease in their brilliancy a line corresponding in position to the brightest of the lines of a nebula strengthened. In December, 1876, this last line was much fainter than the F hydrogen line of the solar spectrum, while in March, 1877, F was much the fainter of the two. The spectrum of the star degraded from its first complex character, so that in the latter part of 1877 it emitted only monochromatic light, corresponding in position to the strongest line in the ordinary nebular spectrum. If the object had then been first observed, any spectroscopist would have pronounced it unhesitatingly to be a gaseous nebula; and it is clear that we have actually witnessed a reversal of the process imagined by Sir William Herschel: that is, a star has actually become a planetary nebula.- Variable Stars. A very remarkable variable star of short period has been discovered by Ceraski. It goes through its changes in about six hours, and alters its color from a blue-white at its brightest, to an orange-red. Prof. Pickering, in discussing the variability of stars of this type, has recently reached the conclusion that they are due to an eclipsing satellite, which in the case of Ceraski's star wholly covers it, but either has a feeble light of its own or else consists itself of a meteoric cloud, which allows about of the light of the primary to pass through. Prof. Pickering considers that, on the other hand, such variables as ẞ Lyra owe their change to the star's own rotation, which presents to us brighter or darker sides.-Photographs of Star Spectra. Most interesting photographs of the spectra of stars and planets have been made by Dr. Huggins of London and Dr. Draper of New York. The former observer now distinguishes them into groups. Those of the first group give 12 very strong lines, all apparently due to hydrogen. The second group give spectra resembling that of our own sun. The third includes the red stars. Spectroscopic observations of the motions of stars in the line of sight have been continued at the observatory of Dr. Huggins, near London, and at that of Greenwich, and are also about being begun by Prof. Young with the great telescope at Princeton, N. J. The late English results show an increasing correctness of estimate, though a great deal is still to be done to attain exactness. About 100 stars have been observed for this purpose at Greenwich, for instance, with the result that while it is now usually safe to pronounce at once whether the star is approaching us or receding from us (so that the direction of the motion is readily known), the amount of this motion remains doubtful within wide limits. Thus at Greenwich, between March and November, 1882, 15 observations of this character were made on Aldebaran, which uniformly concurred as to the fact of its receding, but which varied, as to the estimated rate of this withdrawal, all the way from 5 to 95 m. a sec.-Star Catalogues. The Durchmusterung des nördlichen gestirnten Himmels, the joint work of Argelander and his assistants, Krüger and Schönfeld, embraces all the stars of the first nine magnitudes from the north pole to 2° of south declination. This work was begun in 1852, and at its completion a catalogue of the approximate places of 324,198 stars, with a series of excellent star maps giving the aspect of the northern heavens for 1855, was at the service of astronomers, and has been in the most constant use from that time forward. Argelander's original plan was to carry this Durchmusterung as far as 23° south, so that every star visible in a small comet-seeker should be registered. This plan was abandoned, but Dr. Schönfeld, at the observatory of Bonn, is executing this work. The equinox of 1855 is chosen as the fundamental one; and almost the only changes are the adoption of a telescope of six inches aperture for the work, and a closer discrimination of the magnitudes of the fainter order of stars. Schönfeld has already made 243,000 observations, and astronomers in the northern hemisphere will soon possess an index, as it were, to every star Dr. polar distance. He gives 2,344 stars in this space. The most important contribution of this kind which has appeared since 1843 (the date of the publication of the Uranometria Nova of Argelander) is the Uranometria Argentina of Dr. B. A. Gould (1879). It is the result of nearly ten years' work at the national observatory of the Argentine Republic, and covers the whole southern heavens and 10° of the northern. The magnitudes are on Argelander's scale, but this had to be extended to 7·0 mag., since the naked eye at Cordoba can distinguish stars a full magnitude fainter than at Bonn; that is, stars having only of the light of the minimum visibile of Argelander. The space examined covers about of the whole sky. Within it 10,649 stars were visible under the best circumstances, and more than 46,000 obMany variable stars were discovered. Gould's conclusions as to the situation and shape of the cluster of stars to which our solar system belongs are: "There is in the sky a girdle of bright stars [from first to fourth magnitude], the medial line of which is nearly a great circle inclined to the milky way by about 20°. The grouping of the fixed stars brighter than 4.1 mag. is more symmetric relative to that medial line than to the galactic circle. These facts, together with others, indicate the existence of a small cluster within which our system is eccentrically situated, but which is itself not far from the middle plane of the galaxy. This cluster appears to be of a flattened shape, somewhat bifid, and to consist of somewhat more than 400 stars of magnitudes from 1 to 7, their average magnitude being about 3.6." Dr. C. H. F. Peters, of Hamilton College, Clinton, N. Y., has published the first series of a set of charts of zodiacal stars, of great value, the product of his own observations for over twenty years. The twenty charts already issued contain nearly 40,000 stars. likely to be used in their observations. Prof. | of the sky from the south pole to 70° south O. Stone of the Cincinnati observatory is extending this review still further south. Both his results and those of Schönfeld are as yet unpublished. The astronomical society of Germany, founded in 1865, has instituted a plan for the reobservation of the 324,000 stars from the first to the ninth magnitude (inclusive) found in the Durchmusterung. The observations of Argelander and his assistants were made with a small comet-seeker, and give only approximate positions. The plan of the society contemplates the reobservation of each of these stars by means of large meridian circles, so that accurate places may be available. These stars serve as points of comparison to which are referred the positions of comets, asteroids, &c. By dividing the space to be covered (from the north pole to 92° N. P. D.) into zones, the cooperation of many observa-servations were made, or over 44 for each star. tories is secured, and a prompt publication of the resulting catalogues is made possible. In 1878 the various zones were assigned as follows to the observatories engaging in the work: Nikolayev, -2° to +1°; Albany, 1° to 5°; Leipsic, 5° to 15°; Cambridge, England, 25° to 30°; Leyden, 30° to 35; Lund, 35° to 40°; Bonn, 40° to 50°; Cambridge, United States, 50° to 55°; Helsingfors, 55° to 65°; Christiania, 65° to 70°; Dorpat, 70° to 75°. A catalogue of red stars, with observations of them, has been published by Mr. Birmingham of Ireland: 658 red or reddish-yellow stars are given, none fainter than the tenth magnitude. Secchi has also published a list of 444 red stars. The importance of such lists depends upon the fact that it is found that most red stars are variable, and Birmingham gives it as a result of his observations that for variable red stars the red color in general increases as the star gets fainter, and vice versa.-Uranometry. In 1872 Dr. Heis of Münster published in the Neuer Himmelsatlas the results of 27 years' labor on the stars visible to the naked eye. It is on the same plan, and the scale is the same, as the Uranometria Nova of Argelander. But Heis has extended the scale of magnitudes beyond 60 mag. (where Argelander stopped) to about 6 mag. It may be said, in passing, that under good circumstances he always sees a1 and a' Capricorni divided. More stars are to be found in Heis than in Argelander. The former has 3,507 stars from the first to the sixth magnitude, while the latter has 3,256. Heis has in all 5,471 stars visible to the naked eye under the best circumstacnes at Cologne, Germany; that is, from the north pole to about 130° of north polar distance. One important feature of his work is the accurate delineation of the milky way, both in position and brightness. A similar work was published in 1874 by Dr. Behrmann, who made a voyage during 1866 in the southern hemisphere, and devoted ten months to making a uranometry on Argelander's plan (and with his scale of magnitudes) STEAM HEATING, the warming of the air of inhabited rooms with the heat which is set free on the condensation of steam. The heat stored up in steam, which is released when the steam, without falling in temperature, is condensed to water, is sufficient to raise the temperature of 10 units of water for each unit of weight of steam condensed 50° C., or to raise 10 units of air for each steam unit 210°. The first recorded application of steam to the warming of buildings was made by James Watt in 1784. He constructed an apparatus for warming his study, consisting of a narrow box 34 ft. long with sides of tinned iron placed about an inch apart. The steam was conducted into this heater, which stood on edge, from a boiler, warming the room with the heat liberated on condensation, the condensed water flowing back into the boiler by the same pipe through which the steam ascended into the box. A similar plan was patented in England in 1791 by Hoyle, but failed through a defec tive design. In 1799 Lee devised a heating apparatus in which cast-iron tubes served the double purpose of conveying the steam and supporting the floor. Steam is used both at high and at low pressure in the different appliances invented for warming houses. One advantage which steam heating possesses over heating by hot-water pipes is, that the steam pipes can be made to run in any direction. The only precaution necessary is to provide an outlet for the water of condensation to pass off at every change of level. The condensed water can be used over again in the boiler, and may also be applied to the sanitary uses of distilled water. In the first large building heated by steam, a silk mill in England, the pipes were arranged in the manner which it is desirable to approximate as nearly as possible in all cases, with a gentle and continuous inclination down to the boiler. Though steam heating was regarded with favor and frequently practised in the early part of the century, it fell into neglect after the introduction of the methods of warming with hot-water pipes.The chief difficulty with steam heating is the attention the apparatus requires in order to insure an even supply of steam. When the steam fails, the heaters immediately grow cold. Explosions seldom occur; but there is always more or less danger with careless or inexperienced attendants. Formerly it was considered unsafe to use steam at a higher pressure than 2 lbs. to the square inch in the boiler; but the use of high-pressure steam with engines has become so common that 10 or even 20 lbs. pressure in a steam-heating apparatus is not now considered dangerous, though where a special boiler is used a greater pressure than 5 or 10 lbs. is not desirable. With higher pressure in the boiler and pipes, a smaller radiating surface is required. When low pressure was used, it was considered necessary to have a steam space in the boiler equal to the whole cubic contents of the pipes, so that the pipes would readily fill with steami when the valves were opened. With highpressure steam, smaller boilers are used; yet it is always economical to have as large a boiler surface exposed to the fire as is possible. The boiler and pipes should be so proportioned that the former will evaporate the same quantity of water in a given time as the latter condense. Experiments and calculations show that with a boiler surface of 4 sq. ft. exposed to the direct action of the fire, the surface required to evaporate 1 cub. ft. of water per hour, 182 sq. ft. of pipe will be necessary, the temperature of the room being 60° F., when the pressure is 2 lbs. above the atmosphere; 161 sq. ft. at 10 lbs., 149 sq. ft. at 20 lbs., and 135 sq. ft. at 30 lbs. In common practice, however, a considerably larger boiler surface, perhaps 6 sq. ft., must be allowed to produce the same result. The boiler should be simple in form and easy to clean, larger in proportion to the fire surface than ordinary, and very du rable. The cylindrical boiler with hemispherical ends is one of the strongest forms, and is well adapted for high pressure. Besides the usual fittings of steam boilers, a steam-heating apparatus should be provided with an automatic appliance for injecting water into the boiler. The injector longest in use is the stone-float apparatus, in which a stone is connected with a lever within the boiler, which opens a cock communicating with an elevated cistern. When the water evaporates below a certain quantity, the float falls, opening the cock, and letting in sufficient water to float the stone and close the cock again. With lowpressure apparatus an inverted siphon pipe may be used for the escape of the condensed water. Where steam of high pressure is employed, a number of different contrivances have been devised which allow the water to flow off while preventing the escape of steam. A cock is generally placed at the connection of the pipe with the boiler. When this is opened the steam drives the air in the pipes before it; but unless a means of exit is provided for the imprisoned air whenever the steam is turned on, the pipes in which the air remains will continue cold. When the pipes are laid horizontally, a blowpipe at the further ends allows of the expulsion of the air without difficulty. When the pipes occupy different levels, more complicated contrivances are required; and sometimes it is necessary to employ an air pump. The method of heating buildings by steam can be employed with the greatest economy in premises where steam is used for driving an engine. The boiler should be enlarged to supply steam for warming in the proportion of 1 cub. ft. for every 2,000 cub. ft. of space to be heated; a boiler adapted to an engine of one-horse power can be made large enough to furnish steam for warming a space of 50,000 cub. ft.-The warming of the British houses of parliament is accomplished according to Gurney's modification of the system of Dr. Reid, first tested in the temporary house of commons erected in 1835. Fresh air from without is filtered through screens, and warmed by passing over iron boxes filled with steam, entering the halls through perforated floors, over which porous horsehair cloths are spread to prevent the currents from being felt. In summer the air is cooled by a spray of cold water in the same lower chambers before ascending into the halls. In a series of experiments carried on during the centennial exhibition at Philadelphia in 1876, a very effective method of heating air was shown, which consisted in forcing currents of air over a steam pipe by means of a fan. From the results of trials with an exhaust fan it appeared that while by an ordinary steam coil heat is transferred to the air only at the rate of 0003 of a unit of heat per hour for each degree of difference between the temperature of the steam inside the tubes and the air outsidewhen the air is forced across the tubes at the rate of 39 ft. a second, the heat is imparted to the air at the rate of 87.5 heat units for each degree of difference between that of the temperature in the heater and the external air. This method of heating air is therefore nearly 3,000 times as rapid and efficient. The steam required to drive the fan can be used to furnish the heat. The heater and fan should be so designed that all the steam which is required to furnish the necessary power may be thus utilized. Walker's heater has been much used both for steam and hot water. It consists of a number of small iron blocks, each having square perforations passing through it for the passage of air upward and downward; the blocks being enclosed in an iron box with corresponding perforations, with an inch of space for the steam around each block, which heats the thin metal of which they are composed, while they heat the circulating air. By this arrangement 160 ft. of heating surface can be obtained in a box measuring only 2 cub. ft. The plan of heating hothouses by discharging steam directly into the place containing the plants, the moisture causing them to thrive more luxuriantly, was adopted about the time of the first inventions for house-warming by steam. A simple steam-generating apparatus is made by placing in a furnace a series of concentric heating coils, at the lower end of which the water enters from a reservoir, into which the condensed water returns from the steam pipes. By another arrangement, the pipes in which the steam is generated form the basket grate of the furnace. A sectional boiler is sometimes made by connecting with vertical pipes a series of inclined pipes, communicating at their upper end with the heaters, and receiving the water of condensation at their lower end. A furnace is made in which the steam-generating pipes line the furnace, form the sides of the basket grate below, the bottom of which is given a longitudinal agitation by a ribbed rocking bar, and, winding in a convolved series, fill the space in the furnace above the fire; these horizontal pipes are connected by vertical pipes and headers; the condensed water enters the lower end of the generator from a receiver in front of the furnace. The Holly system of steam heating is a method of heating city buildings from a central source by means of steam conveyed in main pipes through the streets, and conducted to radiators in the houses. This system was first put in practice in Lockport, N. Y., in a form elaborated by Birdsell Holly, a citizen of that place. Other inventors had proposed similar schemes for economical heating. Coleman's, which was never carried into execution, was essentially the same as Holly's. The Holly works in Lockport were put in operation in the autumn of 1877. After settling, by experiments extending over ten years, upon the materials for the conduit pipes, the form of generator, the meters, and all the details of the apparatus, a company was formed with a capital of $25,000, which proposed to heat the dwellings and public halls of the town at a price somewhat below the cost of warming by hot-air furnaces. The second winter 1,000 consumers were supplied, and an aggregate space of about 10,000,000 cub. ft. was heated. The average cost for each consumer was found to be $57.80, counting $18 for interest on the cost of fixtures and that of maintaining them in good order; against $113.75 required to keep a furnace using 10 tons of coal per annum, and $197 required annually to maintain a private steam-heating apparatus consuming 12 tons of coal. The street mains are wrought-iron pipes covered with an inner packing of asbestus, which is enclosed in a jacket of cow-hair felting or some similar non-conducting substance, upon which hard-wood strips are firmly bound with copper wire. The pipes, thus packed and protected, are inserted into holes bored through logs of wood, the holes being large enough to leave an air chamber around them. The logs are laid at an incline over tile drains. A comparison of different sizes of pipes with regard to their capacity for conveying steam shows that the distance to which steam can be carried increases threefold when the diameter of the pipe is doubled: a 13-inch pipe will deliver steam about 1,000 ft. from the boiler; a 3-inch pipe, 3,000 ft.; a 6-inch pipe, 9,000 ft. Steam is served by the Lockport works as far as a mile and a third away from the boiler house. The steam is received from the boiler in a 6-inch main, which branches into two 4-inch, and these again inte two 3-inch mains each. The 3-inch pipes divide into the smallest size of street mains used, 14 inch in diameter. The total length of pipe laid was about five miles. The consumption of coal in the Lockport works the first winter, when only one or two of the six boilers were used at a time, was from 2 to 3 tons a day. With larger boilers and pipes, steam might be conveyed and distributed over an area of four square miles from a single boiler house; although it might be more economical to make the districts to be heated from one station smaller. The isolation of the pipes is remarkably perfect, and the heat which escapes serves a very useful purpose in preventing the hydrants in the streets from freezing up. It is estimated that 10 boilers 16 ft. long and 5 ft. in diameter, with 54 tubes each, will warm an aggregate space of 15,000,000 cub. ft., furnishing steam for heating 1,300 cub. ft. to each square foot of heating surface in the boilers. The loss by condensation in 1,600 ft. of 3-inch pipe, the head of steam being kept at a pressure of 18 lbs., corresponds to the consumption of 9 lbs. of coal an hour; with that length and size of pipe in an ordinary city street, 100 houses can be supplied, making the loss by condensation for each consumer equivalent to 2.16 lbs. of coal a day, With a pressure in the pipes of 60 lbs., or four atmospheres, it required, when the sup ply of steam was cut off, 18 minutes for the pressure to fall to 45 lbs., 28 minutes more for it to sink to 30 lbs., 40 minutes longer for it to decline to 15 lbs., and 54 minutes longer, or 2 hours and 20 minutes altogether, for it to become reduced to the pressure of the atmosphere, or for the four atmospheres of steam to become entirely condensed. In addition to the remarkably efficient service of the coverings and packing used to isolate the pipes, a further economy is effected by a process for reconverting into steam a portion of the water of condensation. Important features in the Holly system are the regulating and measuring apparatus placed in each house, and the distributing and junction chambers which are placed at intervals of 100 ft. or more along the street mains. These junction boxes provide for the longitudinal contraction and expansion of the pipes, affording at the same time a space for the apparatus by which the steam is distributed. The service pipe where it enters the junction box has attached to it a hood into which the water of condensation collects. This is conveyed to a valve in the house, where it is wire-drawn, and in consequence of the reduction of pressure, which is about 50 lbs. to the square inch, a large portion of it is reconverted into steam, which is fed into the radiators. A valve, similar to the slide valve in a high-pressure engine, admits the steam from the street main into a short pipe, which has a similar valve at its other end connecting with the radiators. These valves serve both as regulators and meters. The pressure on both sides of the valves is shown by steam gauges. An indicator registers the consumption of steam in figures which show its value in dollars and cents. The Holly system, after two years of successful operation in Lockport, was introduced in Springfield, Mass., Auburn, N. Y., and other towns. The actual economy over ordinary modes of heating can be arrived at only after it has been at work long enough to test the durability and security of the conducting pipes. The greatest advantage of this method of artificial heating, the reduction in the risk of fire, will be the slowest to obtain practical recognition, owing to the custom of transferring most of the risk to insurance companies, which derive a profit from their guarantee. STEEL, Dephosphorization of. Although the phosphorus is sufficiently eliminated from iron in the puddling process, it is not removed in either the Bessemer or the open-hearth system of manufacturing soft steel. The extreme heat required in the Siemens furnace and Bessemer's converter has been supposed to be the reason why the phosphorus cannot be got rid of. This heat, exceeding that of any other industrial process, necessitated the use of silica as a coating to the furnace or converter, the only substance supposed to be able to resist it. As a consequence, only a limited number of iron ores, containing a minimum of phosphorus, | could be used in the manufacture of Bessemer steel; of the different ores of Great Britain, not more than one eighth were adapted for this purpose. Parry patented a process in 1861 for obtaining pig iron freed from phosphorus and sulphur. Several other methods of purifying iron and steel have since been devised by Bell, Krupp and Bender, Jacobi, Velge, Tessié du Motay, and others; but until the introduction of the Thomas-Gilchrist process the production of Bessemer steel was only possible from a very limited number of ores. The substitution for silica of a basic substance, like lime, in the lining of the furnace, was the means by which these inventors sought to work out the problem of dephosphorization. Snelus and other metallurgists had long before suggested the use of lime, and Siemens had undertaken a series of experiments to that end, but without success. The softness and want of coherence of lime render it a most unmanageable material for furnace linings. Several years were consumed by Gilchrist and Thomas at Blaenavon in experimenting with oxide of iron, alumina, lime, and limestone. They finally succeeded in obtaining, instead of an acid lining like silica, a basic one which was as hard and compact as the best silica bricks. The new bricks, which were strongly basic, were composed of a special kind of magnesian limestone. This was ground, and subjected to intense firing and great shrinkage, passing through chemical changes from which it emerged as hard and firm as silica, and even more infusible. But this was only the beginning of success; for when a pig of phosphoric steel was blown in a furnace lined with these bricks, it was not sufficiently dephosphorized, while the lining itself was so injured that it had to be replaced after four or five blasts. The attempt to assimilate the process of steelmaking to that of puddling was abandoned as impracticable. After long experimentation the right process was evolved, by which the phosphorus is made to combine with the base, and is carried off in the slag. The manufacture of steel by the Thomas-Gilchrist process was commenced in the English district of Cleveland in 1879. The perfected process is complex, and its details have been developed experimentally. To the silica resulting from the oxidation of silicon is supplied lime or magnesian lime, apart from that in the lining, with which it unites at the moment of its formation. The lime and magnesia in the slag must not be less than 40 per cent., and the silica must be less than 20 per cent. When the latter is under 16 per cent., the best results are obtained. The slag from Bessemer and Siemens furnaces ordinarily contains above 40 per cent. of silica, and little lime or none at all, while the cinder contains not over 5 per cent. The quantity of the base added in the new process amounts to from 9 to 14 per cent. of the weight of pig converted, varying with the character of the pig. A portion is added before the metal is |