motives or of the effects of their acts on civilization.

CARPENTER, MATTHEW H., born in Moretown, Vermont, in 1824; died in Washington, February 24, 1881. In 1843 he entered the Military Academy at West Point, where he remained two years. He then went to Boston and studied law with Rufus Choate, and was admitted to the bar. In 1848 he removed to Wisconsin, and entered the practice of his profession, in which he soon became eminent for his legal ability and brilliant talent, which won him high reputation in the Supreme Court of the United States even before his entrance into public life. Not until after the war did Mr. Carpenter take an active part in politics. Before and during the war he was a Democrat; but, when the leading men took sides on that issue, he became a Union inan. When, at the close of the contest, he espoused Republicanism, his generally recognized ability commanded for him the active support of that party in Wisconsin, and in 1869 he was elected to the United States Senate in place of Senator Doolittle.

Mr. Carpenter served in the Senate from May 4, 1869, until March 3, 1875, and occasionally showed great power as a lawyer and debater, but lacked those qualities necessary to make a public man understand public sentiment. He belonged to that class of brilliant politicians who so strongly influenced the proceedings of Congress from 1869 to 1875, and of which General Butler was a representative man. About this time Mr. Carpenter was the victim of malicious slanders, but he was able to prove to general satisfaction that they were groundless. In 1874 Wisconsin Republicans, like the party elsewhere, were suffering from the injudicious action of Congress upon the salary bill and like matters, and the feeling against the railroad corporations was also a distracting element. The party had, however, a majority in the Legislature, but a considerable portion of it was made up of Independents. Mr. Carpenter received the caucus nomination for Senator, but the independent minority refused to vote for him. After a protracted struggle, the Democrats joined the independent Republicans and elected Mr. Cameron. Mr. Carpenter accepted his defeat, vouched for the Republicanism of his successor, and retired to his extensive law practice, taking little interest in political affairs. During the contest over the presidential succession of 1876-'77, Mr. Carpenter appeared before the Electoral Commission as one of the Tilden counsel, and made an argument in his behalf. The Legislature of Wisconsin, which met in January, 1879, was called upon to choose a successor to Senator Howe. The contest between Messrs. Howe, Keyes, and others was a bitter one, and finally Mr. Carpenter was presented as a compromise candidate. He had been approved on financial questions, and his superior talents rose paramount over all the opposition formerly urged.

His election gave general satisfaction to the Republicans. He took his seat March 4, 1879. Among several speeches which he addressed to the Senate, all remarkable for their ability, that against the Fitz John Porter bill is regarded as his finest effort. His course in politics during his last term in office was much more independent than previously, and as a lawyer he had few equals in Washington, where most of his later years were spent.

CENSUS OF THE UNITED STATES. (See UNITED STATES CENSUS.)

CHEMISTRY. The president of the Chemical Section of the British Association, Professor A.W.Williamson, made the growth of the atomic theory during the last fifty years the subject of his opening address at the last year's meeting, maintaining that its general validity had been confirmed by the tests of experimental application to which it had been rigorously subjected. The binary or dualistic theory of combination, and the doctrine of multiple proportions which were formerly connected with it, and which seemed to be satisfactorily applicable to the simpler compounds and the salts, broke down when chemists came to deal with double compounds which were not saline in character, and with the carbon compounds; and it became necessary to find other methods of accounting for the phenomena of chemical combinations. As the investigations were continued with reference to this object, the idea of substitution came to be more and more used in the place of that of mere additive combination. Elementary chemical reactions which, according to the binary theory, were supposed to consist of dualistic processes, involving sometimes the assumption of forces (like predisposing affinity) of a purely metaphysical character, were explained as consisting of atomic displacements, or interchanges of a kind well known to be of common occurrence. Many products of the combination of known molecules have been found to be formed by processes of double decomposition, so that each molecule of such products is built up partly of atoms derived from one of the materials, partly of atoms from the other. In the organic compounds, many of the molecules having a very complex structure have been found to undergo in their reactions very simple changes, of the same kind as those which mineral compounds undergo. Families of organic compounds have been discovered whose members are connected by close analogy of constitution and properties, each of them forming a series, each term of which is a compound whose molecule contains one atom of carbon and two atoms of hydrogen more than the previous one. Our knowledge of molecules has undergone a most remarkable and important extension by the discovery that various elements in what we are accustomed to consider the free state, really consist of molecules containing like atoms combined with one another. Those marvelous varieties of matter called isomeric compounds have found

their natural explanation in differences of the respective arrangement of like atoms. The term "equivalent" was introduced to indicate the proportional weights of analogous substances which were found to be of equal value in their chemical effects. Tables of the equivalent weights of acids were made, representing the proportions of the several substances that were found to be of equal value in neutralizing a fixed quantity of a certain base; and similar tables were made for the bases, as well as for the elements themselves. But little attention was paid under the dualistic system to the essential difference between atomic weights and equivalent weights; but under the later developments of the theory of the functions of atoms, it has become necessary to study the relation of equivalence between elementary atoms, instead of studying them from the point of view of elements divisible in any proportion. From this has sprung the division of the elements into classes consisting respectively of equivalent atoms known as monads, dyads, triads, tetrads, etc., the character of which is well represented in the four typical compounds, Cl H, OH2, NH3, CH4, where chlorine appears as a monad, oxygen as a dyad, nitrogen as a triad, and carbon as a tetrad. This has probably been one of the most important steps yet made in the development of the atomic theory, and has been seen to correspond in so clear and striking a manner with a vast number of well-known properties and reactions of compounds as to deserve and acquire the confident trust of chemists. Another great step has recently been made which may be destined to lead to most important results. It has been discovered that if we arrange the elements in the empirical order of their respective atomic weights, beginning with hydrogen, and proceeding thence step by step up to the heaviest atom, we shall have before us a natural series with periodically recurrent changes in the chemical and physical functions of its members. Of course the series is still imperfect, and exhibits gaps and irregularities; but some of the gaps have been filled up by the discovery of elements possessing the anticipated properties, inducing the hope that the others may be filled. The generalization affords a brilliant addition to the previous corroborations of the reality of the units of matter which chemists have discovered. But little account has as yet been taken of atomic motions; but it has been proved that the heat of combination affords a measure of its force, and we know that, in giving off heat, particles of matter undergo a diminution of the velocity of their motion. The force of chemical combination is evidently a function of atomic motion, but a vast amount of work will be required to develop the atomic theory to the point of explaining the force of chemical action in precise terms of such motion.

ATOMIC AND MOLECULAR WEIGHTS.- Variations in Atomic Value.-Professor A. W. Will

iamson regards the opinion that atomic values are invariable, or are variable only within particularly defined limits, as an error. He remarked in a recent address that he had been frequently struck by the fact that two theories, believed at one time to be conflicting, had been shown by the progress of study to be both true. Such was the case with the rival theories, one of which represented molecules as constructed after the pattern of three or four types, while the other viewed them as containing complex groups called radicles. Opposition existed at one time between those who made use of atomic weights and those who employed equivalent weights; and the most important step that has of late been taken is the introduction of the notion of equivalence into the atomic theory. An inspection of the series of hydrogen compounds with chlorine, oxygen, nitrogen, and carbon, will show that the atom of chlorine, which combines with a single atom of hydrogen, has a different value from the atoms of oxygen, nitrogen, and carbon, which combine respectively with two, three, and four atoms of hydrogen. Hence, nitrogen and other elements of like equivalence are called trivalent or triads, while other elements are classed, according to the exponents of their equivalence in groups, as monads, dyads, pentads, etc. Kekulé still holds that an element can belong to only one of these groups; that nitrogen, for instance, is trivalent only, and that in sal-ammoniac, where it is combined with five other atoms, instead of being pentivalent, it is a molecular compound of two chemical compounds-ammonia and hydric chloride; and that the atoms of constituent molecules and the molecules themselves were held together by different forces, one being molecular, the other chemical. We have, however, no grounds for assuming a difference between chemical and physical forces, and Kekulé's theory is no longer tenable. The theory commonly in vogue is that atoms vary in value within certain narrow limits; that nitrogen, for instance, is either trivalent or pentivalent. Professor Williamson recognizes no limitation to atomic value; but he knows that many elements have atomic values greater than those commonly assumed. The character of the atoms often materially affects the result. Thus gold can not combine with more than three atoms of chlorine alone, but it can take up an additional atom of chilorine if supplied with an atom of sodium. In this way we get the common double chloride of gold and sodium, NaAuCl, in which the gold is pentivalent. We are not to consider the sodium as being here combined with the gold as such, but as combined with the whole group. Temperature, also, has an influence upon the atomic value of elements, a rise of temperature tending to diminish it.

Molecular Weight of Hydrofluoric Acid.Professor J. W. Mallet has made some studies of the atomic weight of hydrofluoric acid, with a view to finding an explanation of some

peculiar differences in the behavior of fluorine in entering into combination with other elements. The analogies of fluorine with the halogens on the one hand, and with oxygen on the other hand, have often been remarked upon. The compounds of fluorine generally bear resemblance to compounds of chlorine, but some striking differences in the character of these compounds have also forced themselves upon the attention; and the tendency of the fluorides to the formation of double salts, with formulas analogous to those of oxygen compounds, and the formation of salts including both oxygen and fluorine, has suggested that, some close natural relation may exist between these elements themselves. There has, therefore, been ground for questioning whether fluorine should be classed with chlorine among the monad elements, with the formula IIF to represent hydrofluoric acid, and assigned an atomic weight of 19, or with oxygen among the dyads, with the formula HF for hydrofluoric acid, and an atomic weight of 38. Professor Mallet's experiments bore a special reference to this question. The result was such as to justify the assumption that at the temperature of 30° centigrade the molecule of hydrofluoric-acid vapor should be represented, not as HF, but as H2F2, while at higher temperatures dissociation takes place, leading to the production of diatomic molecules of HF. The structure of the molecule of double weight, HaF2, may be regarded as resulting from fluorine behaving not only as a monad, but also as a triad, and acting in double atoms like those of nitrogen in the di-azotic compounds. In such a condition the element presents a pseudo-dyad character, and becomes capable of replacing oxygen and of manifesting the linking function of that element. This assumption, supported by the experimental evidence brought forward by Professor Mallet, may serve conveniently to explain the composition of a number of fluorine compounds, whose formulas are difficult to write in a satisfactory way if fluorine be considered exclusively monad.

Atomic Weight of Platinum.-The group of metals embracing osmium, iridium, and platinum has until recently exhibited a series of irregularities in that their atomic weights did not manifest those relations to each other which their properties, in connection with Meyer and Mendelejeff's theory of classification, indicated they should bear. Dr. K. Seubert, two years ago, undertook the revision of the atomic weight of iridium, and fixed it at 192644, putting it below that of platinum. He has since fixed the atomic weight of platinum at 194 177, giving it the place above that of iridium and below that of gold, which the theory requires it should occupy, while the previous estimation of its atomic weight made it above that of gold. The ascending series, iridium, platinum, gold, is now, as to those three metals, agreeable to theory; but osmium still occupies an anomalous position, its re

ceived atomic weight, 1985, being above that of gold, while the theory requires that it should be below that of iridium.

Molecular Weights of Decipium and Samarium.-M. Delafontaine, in 1878, described an earth having a molecular weight of about 122, which he had obtained from samarskite, and which he called decipia, regarding it as an oxide of a new metal, decipium. He has continued his studies of this substance, and has been brought to regard it as a mixture of two oxides, one of them having a molecular weight of about 130, and the other a lower molecular weight. The former substance gives no absorption spectrum, while the second gives the spectrum which M. Delafontaine described in 1878 as that of decipia. M. Lecoq has also announced the discovery of a new metal in samarskite, corresponding with the second substance detected by M. Delafontaine, to which the latter proposes to give the name of samarium. The molecular weight of its oxide is believed to be less than 117. Samaria appears to be identical with the earth Y3, having a molecular weight of 115, which M. Marignac has found in samarskite, while that chemist's Ya, having a molecular weight of 120-5, may be supposed to be a mixture of decipia and terbia.

Atomic Weight of Aluminum.-Professor J. W. Mallet, considering that the estimation of the atomic weight of aluminum was resting on an insufficient basis of accurate experiment, has pursued, during three years, a course of experiments for the revision of the determination, in which he has kept in view the principles-1. That each process used should be as simple as possible, and should involve as little as possible of known liability to error; 2. That different and independent processes should be resorted to as the means of checking each other's results; 3. That each process should be carried out with quantities of material differing considerably from each other in successive experiments; 4. That only such other atomic weights should be involved as may be counted, among those already known, with the nearest approach to accuracy. Three series of experiments were conducted, of which the first series was based on the purification of ammonium alum; the second on the preparation and purification of aluminum bromide; and the third on the preparation and application of pure metallic aluminum. The mean result of the twenty-five experiments which were regarded as the more accurate of the thirty that were made, gives the atomic weight of aluminum as 27:02. This is believed by Professor Mallet to bear in favor of Prout's law, which assumes that all the atomic weights are multiples of that of hydrogen.

Atomic Weight of Cadmium.-Mr. Oliver W. Huntington, under the direction of Professor J. P. Cooke, of Harvard College, has made a revisionary determination of the atomic weight of cadmium. He used a pure bromide of cadmium, specially prepared for the pur

pose, and the bromide of silver, likewise specially prepared, for comparison. The mean of a series of eight experiments gave 112.31 as the atomic weight of the metal. This determination is regarded as bearing against the validity of the hypothesis of Prout, that all atomic weights are multiples of that of hydrogen.

NEW PROCESSES.—Mr. Alfred H. Allen has indicated some valuable simple tests for the presence of hydrocarbon oils as adulterations in animal and vegetable oils. The methods for the detection of these oils are based on the density of the sample, the lower flashing and boiling points, the fluorescent character of the oils produced from petroleum, bituminous shale, and rosin, and the incomplete saponification of the oil by alkalies. The taste of the oil and its odor on being heated are also useful indications. If undoubtedly fluorescent, an oil certainly contains a mixture of some hydrocarbon, but the converse is not strictly true, as the fluorescence of some varieties of mineral oil can be destroyed by chemical treatment, and in other cases fluorescence is wanting. The greater number of hydrocarbon oils employed for lubricating purposes are, however, strongly fluorescent, and the remainder usually become so on treatment with an equal measure of strong sulphuric acid. If strongly marked, the fluorescence of a hydrocarbon oil may be observed in presence of a very large proportion of fixed oil, but, if any doubt exist, the hydrocarbon oil may be isolated. The fluorescence may be seen by holding a test-tube filled with the oil in a vertical position in front of a window, when a bluish "bloom" will be perceived on looking at the sides of the tube from above. A glass rod dipped in the oil and laid on a table in front of the window so that the oily end shall be projected in the view against the dark background of the floor, or a piece of black marble or smoked glass rubbed with a streak of oil and held horizontally before a window, will make a very slight fluorescence, readily perceptible. Turbid oil should be filtered, to get out the minute particles that might, by reflection, give an appearance of fluorescence. Dilution with ether, to which a little mineral oil imparts a strong blue fluorescence, gives an excellent test. The hydrocarbon oil may be driven off by heating it if its boiling-point is comparatively low, but may be better removed and the quantity of it measured by saponifying it, and washing the solution of the soap with ether. The hydrocarbon may, in this case, be recovered pure by separating the ethereal layer and evaporating it at or below a steam-heat. A good alkaline preparation for this purpose can be made by dissolving caustic potash in methylated spirit. The washing with ether should be repeated several times. The etherprocess is, however, not applicable to spermaceti and the waxes, on account of the large quantities so small in the other fats that it need not be taken into account-of matter they contain that is not acted upon by the alkalies

but is dissolved by ether. The nature of the hydrocarbon oil may be determined after it has been isolated, by observing its density, taste, smell, behavior with acids, and other qualities.

Professor G. Lunge, of Zürich, has perfected and described a simple and inexpensive process for procuring pure naphthalen that will not discolor. Presuming that the discoloration of naphthalen is analogous to that of phenol, he sought to remove the agent which caused it by oxidation. For this purpose he added an oxidizing agent in the ordinary chemical washing .of naphthalen, using manganese dioxide, with complete success. Other oxidizing agents might be substituted for manganese dioxide, but a cheaper one can hardly be obtained. Naphthalen prepared by this process has kept its pure white color much longer than the "chemically pure "naphthalen made by the secret process of the manufacturers.

A patent has been taken out by M. Closson, of Paris, for a cheap and expeditious method of obtaining magnesia from magnesium chloride. The crude lye of magnesium chloride is treated with burned dolomite, or magnesian limestone, when the chlorine of the lye combines with the lime of the dolomite, so that if the latter is pure a magnesia of from 98 to 99 per cent standard can be easily made on a large scale. The magnesia bricks prepared by this process at Leopoldshall resist even the flame of the oxyhydrogen-blast. The cost is fifteen shillings a ton. Sulphate of lime is obtained as a by-product of the process through the use of calcium chloride to remove the magnesium sulphate that is present in the magnesium chloride, and is used by paper-makers under the name of pearl-hardening. The value of the new process in its bearing on the manufacture of fire-proof furnace-linings, crucibles, etc., is very great.

Herr A. Wagner recommends the following process, which has proved very satisfactory for the limestone waters of Munich, for the determination of the organic matter in water. After the determination of all the other constituents of the water, he evaporates suitable quantities to dryness and separates the dry residue by means of distilled water into an insoluble and a soluble portion, the latter of which contains the chief bulk of the organic bodies. In the insoluble portion, which in the waters he has to deal with consists almost entirely of calcium and magnesium carbonates, he determines the organic matter by igniting a dried specimen in a platinum crucible and treating subsequently with ammonium carbonate in the customary manner. The portion soluble in water, if nitrates are absent or are present only in a quantity too small to be determined, is dried after evaporation in a platinum capsule, weighed, heated to a very low redness, and weighed again. If nitrates are present in larger quantities, so that the existing organic matter would not suffice to con

vert the nitrates into carbonates, he adds to the soluble portion, after drying and weighing, a little pure solution of sugar, evaporates to dryness, and heats the platinum capsule gradually and by piecemeal with a very small gasflame, so that no deflagration may happen. After prolonged but very gentle ignition, the sugar-charcoal is found burned away, when the residue is moistened with water containing carbonic acid, and weighed again after drying. The loss of weight in this case expresses the weight of the organic substances and the difference between the equivalent of the nitric acid which was present and of the carbonic acid which has taken its place. This difference can be easily calculated from the quantity of the nitrates as previously ascertained, and must be deducted. This process is not absolutely accurate, but Herr Wagner considers it more certain than others. Herr Wagner calls attention to the necessity, in experiments for determining the solid residues, of protecting the platinum or porcelain capsule in which the water is treated against the accumulation of a deposit from the gas-flame, through which a liability to error in weight is incurred. For this purpose he uses a thin sheet of platinum, instead of the ordinary wire gauze, between the capsule and the flame.

Mr. Thomas Moore has published the following new process for the separation of nickel and cobalt from iron: Having removed any excess of free acid by evaporation, and dissolving the residue in water, add to the solution a sufficient quantity of ammonic sulphate to form a double sulphate with the nickel and cobalt present. Dilute to about 150 c. c., and add a rather large excess of oxalic acid, and stir well. In case a precipitate form, more ammonic sulphate should be added until a clear solution is obtained. Add ammonic hydroxide in considerable excess; stir, heat gently for a few minutes, and filter; wash well with water containing ammonia; or dilute to about 500 c. c., and, after allowing the precipitate to settle, withdraw a given portion of the clear upper stratum of liquid. This, after a further addition of ammonic sulphate, to lessen the resistance to the electric current, is ready for electrolysis or any other method of estimating the nickel or cobalt.

Messrs. R. H. Chittenden and H. H. Donaldson, of the Sheffield Laboratory, Yale College, describe a process for the detection and determination of arsenic in organic bodies, which they recommend as very accurate, delicate, and simple. It is based upon Gauthier's process, and somewhat resembles it, but requires for reagents only nitric acid, sulphuric acid, and zinc. The organic matter is destroyed by successive oxidations with nitric and sulphuric acids, as in Gauthier's method, but at a much lower temperature. The suspected matter is then treated for fifteen minutes at 200° C., and allowed to cool, when a hard, carbonaceous residue, free from nitric acid, and containing

the arsenic as arsenious acid, is formed. This is extracted with water till it has been made to give up its arsenic, and the reddish-brown fluid containing some organic matter is evaporated to dryness. The residue is dissolved at a gentle heat with a definite quantity of dilute sulphuric acid, and introduced to a Marsh's apparatus for the decomposition of arseniuretted hydrogen by artificial heat, to which a Bunsen wash-bottle and a device for graduating the admission of the fluid have been added. The resultant gas having been dried in a chloride of calcium tube, is passed through a red-hot glass tube. Not a trace of arsenic passes by if the cooled tube is of proper length. The apparatus is then filled with hydrogen generated by the sulphuric acid-zinc process, and the glass tube, having been heated to redness, the arsenical solution in concentrated form is mixed with sulphuric acid, and the mixture is slowly passed into a separating funnel; then more and stronger acid is added, and the heat is kept up till the decomposition is wholly effected. The arsenic being collected in the form of a mirror of metal, the tube is cut, at a safe distance from the mirror, and weighed. The arsenic is removed by heating, and the tube is weighed again, when the difference gives the amount of metallic arsenic. The method is capable of detecting as little as the one thousandth of a milligramme of the metal. In organic matters the experimenters have detected a millionth of a gramme in urine and in an extract from beef. The experiments are claimed to show that the presence of organic matter in considerable quantity does not interfere with the recovery of the entire amount of arsenic.

VEGETABLE ANALYSIS.-Professor Henry B. Parsons, of Washington, D. C., has described a method for the more accurate analysis of plants. His apparatus includes a worm of block-tin pipe, suitably connected with a glass percolator, within which is suspended a smaller tube, having a bottom of filtering-paper and fine, washed linen. The weighed sample of the finely-powdered herb is placed within this tube for extraction.

The solvent is introduced in a glass flask, tightly fitted to the outer percolator, and is volatilized by the application of heat through a water-bath. A tared filter, prepared by allowing fine asbestus, held in water, to settle on the perforated bottom of a platinum crucible, is also provided and connected with the receiving-vessel, while this in turn is connected with a Bunsen's pump. The air-dried specimen of the plant to be analyzed should be ground or beaten till all the particles will pass through a sieve having from forty to sixty meshes to the linear inch. A part of this should be further pulverized till it will pass through a sieve having from eighty to one hundred meshes to the linear inch. The finer part of the sample is employed in the immediate analysis, while the coarser part is reserved for the separation of those proximate principles