NEWS

P4O6S4+6H20=4HPO3+4H2S. It is also soluble in carbon bisulphide, from which it is again obtained upon evaporation in tetragonal prisms.

Sulphur trioxide is reduced by phosphorous oxide, the sole products being sulphur dioxide and phosphoric anhydride.

Sulphuric acid dropped upon phosphorous oxide occasions great rise of temperature; with quantities of a grm. and upwards incandescence occurs. Sulphur dioxide is liberated, and phosphoric acid formed.

Sulphur chloride, S2Cl2, also acts with violence, forming phosphoryl and thiophosphoryl chlorides, sulphur and sulphur dioxide

P1O6+6S2Cl2=2POCI3+2PSC13+2SO2+8S. Ammonia gas acts slowly in the cold, but with incan: descence on warming to 30°. When ammonia is led over the oxide dissolved in ether, a white solid product is deposited, consisting of the diamide of phosphorous acid, P(OH)(NH2)2, and a small quantity of the corre sponding ammonium salt

PO6+8NH3-3P(OH) (NH2)2+ P(OH)2(ONH4)2 Phosphorous diamide is a white powder which is instantly dissolved by water with sufficient rise of temperature to induce incandescence. Treated with dilute hydrogen chloride solution, it evolves pure, non-spontaneously inflammable, phosphoretted hydrogen, formed by the decomposition at the high temperature of the reaction of the phosphorous acid first formed

POH) (NH2)2+2HCl+2H2O=2NH4Cl + P(OH)3. Nitrogen peroxide vapour oxidises phosphorous oxide in the cold to phosphoric anhydride, being reduced to nitrogen tri- or di-oxide.

Phosphorus pentachloride acts with great energy on phosphorous oxide. The action may be controlled by cooling with ice. A liquid mixture of phosphorus and phosphoryl trichlorides is produced

P406+6PCl5=6POC13+4PC13. Phosphorus trichloride and phosphorous oxide only interact at a temperature near the boiling-point (173°) of the latter, and in a closed vessel; under these circumstances no phosphoryl trichloride is formed, but a solid mixture of pentachloride and pentoxide of phosphorus, together with amorphous phosphorus.

Hydrogen, phosphoretted hydrogen, carbon monoxide, carbon dioxide, sulphur dioxide. nitrogen, nitric oxide, cyanogen and ethylene, have apparently no action upon either cold or warm phosphorous oxide.

DISCUSSION.

Prof. RAMSAY inquired on what grounds the authors

came to the conclusion that the amide described was a

phosphorous and not a phosphoric compound, such as is represented by the formula OPH(NH2)2 corresponding to the formula OPH(OH)2 for phosphorous acid, which latter he thought was undoubtedly a phosphoric compound. The low boiling-point and other properties of the sulphoxide described seemed to lend additional support to the view that phosphoric oxide was a compound of high molecular weight.

Dr. ARMSTRONG remarked that the conventional view that the affinity of phosphorus for oxygen was very great was somewhat disturbed by the observations now described, showing that in many cases the phosphorus and oxygen in phosphorous oxide were readily separated; it would seem that the extreme stability of phosphoric oxide was probably conditioned by peculiarity of structure.

The authors, in their reply, said that they had no special experimental proof of the constitution of the diamide formed by the action of ammonia; Prof. THORPE, however, expressed the opinion that the balance of evidence was in favour of the view that phosphorous acid is P(OH)3.

1

In the first communication to the Society by one of the authors, in conjunction with Mr. H. H. Robinson (Trans., 1890, 38), it was concluded that frangulin had the composition C22H22O9. This conclusion was based on analytical results alone, and it was, in fact, pointed out that the percentage yield of emodin obtained on hydrolysing frangulin agreed better with Schwabe's formula, C21H2009.

The authors have now prepared a larger quantity of frangulin, and have succeeded in obtaining it more nearly in a state of purity than was previously possible. They find, however, that crude frangulin contains a substance isomeric with emodin, which clings to it very persistently, and which it is very difficult to completely remove from previous experimenters to the presence of this substance. the frangulin; they ascribe the conflicting statements of Schwabe's formula, C21H 2009. They have succeeded in proving the correctness of Their conclusions are based not only on their more recent analyses of frangulin, but also, and in fact mainly, on the results obtained from the hydrolysis of frangulin. The two products of the hydrolysis are emodin, C15H1005. as shown in the first paper, and rhamnose, C6H12O5. The latter substance was obtained in a crystalline form, and was identified by its chemical and physical properties and by the properties of its osazone. The percentage yield of emodin shows that one molecular proportion of frangulin yields, one molecular proportion of emodin, and the difference represents one molecular proportion of rhamnose. It is thus possible to build up the formula of frangulin from its constituents, from the analyses of the glucoside, the conclusion is justiand since the result agrees with the formula deduced fied that frangulin has the composition C21H2009, and that its hydrolysis takes place in accordance with the equation: C2H20O9 + H2O=C15H10O5+C6H1205, which was given by Schwabe as highly probable.

The substance which has been mentioned as clinging so persistently to the frangulin has been isolated, and is found to have the same percentage composition as emodin. It melts at 202-203°, and is, without doubt, the same substance which Schwabe found and described as melting at 199. It differs from emodin also in that it crystallises in golden-yellow needles, in the greater readiness with which it sublimes, and in its reaction with alkalis. It is probably an isomeric trihydroxymethylanthraquinone.

70. "The Structure and Chemistry of Flames." By ARTHUR SMITHELLS, B.Sc., and HARRY INGLE, B.Sc., Yorkshire College, Leeds.

The authors have been engaged for twelve months in known hydrocarbons, and are still continuing their experiinvestigating the chemistry of flames produced by burning ments. The publication of their results in the present form is consequent on the appearance of a paper by N. Teclu (Fourn. Prakt. Chemie, xliv., 246), who describes authors' inquiry. the phenomenon which served as a starting point in the

the metal tube of a Bunsen burner, so as to form a wider If a long glass tube be fitted by means of a cork over continuation of it, the flame can be caused to burn at the top of the glass tube. When the gas is turned slowly off the flame becomes smaller, and develops a sharply defined inner cone of a greenish colour; this cone ultimately becomes almost a flat disc of flame and enters the glass tube. It will, as a rule, descend at a rapid rate for some distance, then begin to oscillate and finally either detonate and light the gas at the bottom of the metal tube or else

cone-that is, to prevent its re-ascent. It is also possible

to effect this by narrowing the bore of the glass tube at one point, so as to diminish the rate of inflammation, i.e., to prevent the descent of the inner cone past that point. In this way it is possible to separate the two hollow cones of combustion which constitute the Bunsen flame, and to keep them any distance apart for any length of time. This permits of the aspiration of the gases from the space between the cones without any chance of admixture of outside air or of products of combustion from the upper

cone.

The apparatus used by the authors in most of their experiments consisted of two glass tubes, one of which slides very easily within the other. The inner tube (c), which is the longer one, is united to the outer one by an india-rubber collar (a), through which it slides freely, and the two tubes are kept coaxial by a ring of asbestos packing (b). The projecting end of the inner tube may be fitted to a Bunsen burner, but the authors have usually led separate supplies of gas and air into the apparatus by a T-tube, instead of using a Bunsen burner, in order to

d

have a better control of the flame. With this apparatus a non-luminous flame is easily obtained, and the two cones can be separated in two ways. If the apparatus is arranged so that the flame is formed at the orifice of the wider tube (d), and the orifice of the narrower one (e) is 8 or 10 c.m. below it, on increasing the air supply the inner cone of flame will ultimately descend and rest upon the orifice of the inner tube. If, on the other hand, the inner tube be made to project beyond the outer one and the non-luminous flame be formed on it, then, if there be a sufficient air supply, on sliding up the outer tube it will, as it passes the flame, cleanly detach and carry up the outer cone, leaving the inner one still burning on the

flame is obtained and divided into two cones. In the case of liquid hydrocarbons, the lower cone of flame usually appears to be divided by dark spaces into several petallike divisions which are in rapid rotation. In the case of benzene vapour the following sequence of appearances is presented :-Starting with the orifice of the inner tube, 8 or 10 c.m. below that of the outer one, a luminous smoky flame is first obtained at the latter; as air is gradually added the flame becomes less and less luminous, and an inner cone begins to develop, but before this has become non-luminous it descends to the inner tube; more air makes both it and the other cone non-luminous, and this state may be maintained. If now somewhat less air is supplied, a luminous streak appears at the tip of the inner cone, and passes right up and through the tip of the upper cone. If more air is supplied, the upper cone of the flame begins to disappear, and only the upper part of it remains; this also gradually fades away, and then there is only the lower cone left. Still more air produces a visible effect on the inner cone, the colour changing and the combusbustion becoming less intense until the cone rises from its seat, passes upwards and disappears. There is thus a gradual transition from the richly luminous flame to one consisting of a simple pale blue cone just on the point of extinction through excess of air.

The hydrocarbons examined by the authors were ethy. lene, methane, pentare, heptane, and benzene. Coal gas was also used. The gases from the regions between the two cones of flame were analysed in all these cases volumetrically or gravimetrically. The following are some of the results obtained:

These results, which are obtained in the preliminary survey, are not quite accurate, owing to the impurity of the hydrocarbons and certain difficulties which are described in the paper. They show, however, that the products of combustion of the first cone are essentially CO2, H2O, CO, and H2, and that the second zone is due to the combustion of the CO and H2 with the external air.

The results are in harmony with the conclusions of Blochmann, obtained indirectly, and with the not generally known work of Dalton on the explosion of methane and ethylene with oxygen in quantities insufficient for complete combustion, which was repeated in 1861 by Kersten.

The authors point out (i.) that carbon, according to Baker's experiments, even in excess of oxygen, burns preferentially to CO, and not to CO2; (ii.) that the heat of combustion of gaseous carbon to CO is probably greater than that of hydrogen to H2O; (iii.) that, according to Dalton, CH, when burnt with its own volume of oxygen gives products represented in the equation

CH, + O2 = CO + H2O + H2;

and they conclude that this equation represents the character of the change first taking place in the inner But as the two substances CO and H2O act upon one another

cone.

(CO + H2O CO2 + H2),

the case is one of reversible change, and four products

will result, viz., CO2, H2O, CO, and H2.

The conditions of equilibrium of this system, according to Dixon, are expressed by the coefficientCO × H,O = 4'0. CO2 × H2

This is subject to certain conditions of temperature and dilution. The authors in their most reliable experiments (viz., the gravimetric ones), with ethylene and coal gas, get numbers not greatly differing from 4; but they are still engaged in studying this question.

The authors have succeeded in dividing into two cones flames produced by admixture of air with cyanogen, sulphuretted hydrogen, carbon bisulphide, and decomposed ammonia (ie., N2+ 3H2). The products of the inner cone in the case of cyanogen were found in one experi

ment to consist of CO and CO2 in the proportion of 2 vols. of the former to one vol. of the latter.

Professor Smithells is continuing the experiments with a view of elucidating the following points:

(i.) The influence of differences of diameter of the tubes and rates of efflux on the fractional co.nbustion. (ii.) The exact composition of the interconal gases in the case of hydrocarbons, and also of mixtures of CO+H2, so as to ascertain if, and in what way, the coefficient

varies with the composition of the gases and other

conditions. (iii.) The composition of the interconal gases from hydrocarbon flames whilst carbon is being liberated, so as to ascertain whether the luminosity of flame is due to simple decomposition of hydrocarbons by heat, to preferential combustion of hydrogen, to partial decomposition, or to other change. (iv.) The exact nature of the flame of cyanogen, so as to ascertain what governs the proportions of CO and CO2 formed in the inner cone. (v.) The manner in which the partition of oxygen takes place in the inner cone between C and H, H and S, C and S, so as to obtain information as to the affinities of C, H, and S for oxygen. (vi.) The spectroscopic appearances of the flames. DISCUSSION.

Professor THORPE after expressing regret that time did not permit of the discussion of the numerous interesting questions which had been raised, referred to Professor Smithell's remark, that the books failed to notice the fact that carbonic oxide was produced on partial combustion of methane. He pointed out that Mr. Thomas, several years ago, made a special study of this question, having been led to do so by the observation that in cases of undoubted marsh gas explosions in coal mines the men killed often exhibited an appearance suggestive of carbonic oxide poisoning, Mr. Thomes found that carbonic oxide and hydrogen were regular products of the incomplete combustion of marsh gas. (Cf." Coal Mine Gases and Ventilation," by J. W. Thomas. Longmans and Co., 1878. Also Iron, 1875).

(To be continued.)

a shape as to give a nearly closed circuit of small "magnetic resistance." The pattern now described consists of a a steel rod I inch diameter and about 2 inches long, with a cast-iron disc 4 inches diameter and thick, fixed at one end; the other end is fitted in a hemispherical iron shell which surrounds the bar and comes flush with the upper sur:ace of the disc. An annular air space, less than wide, is left between the cylindrical surface of the disc and the inside of the shell; and when the bar is magnetised a strong magnetic field exists in this space. To use this field for producing electro-magnetic impulses, a coil of wire is wound, in a shallow groove, on a brass tube which can slide easily through the annular space, thus cutting all The tube is allowed to fall by its own weight, the lines. a neat trigger arrangement being provided for effecting its release.

The instrument exhibited had go turns of wire in the coil, and the total magnetic flux across the air space was pulse is therefore obtainable even through resistances as about 30,000 C.G.S. lines. A large electro-magnetic imgreat as 10,000 ohms. Tests of three instruments show that there has been practically no magnetic decay in seven months. The author therefore considers them satisfactory, and is prepared to supply them as magnetic standards. To facilitate calculations the number of lines will be adjusted to a convenient number, say, 20,000 or 25,000.

Several uses to which the instruments are well suited are mentioned in the paper, and a simple way of determining permeability by the magnetometer method is described.

ment was inappropriate, for it really gave a constant im. Mr. BLAKESLEY thought the name given to the instru pulsive E.M.F.

Dr. SUMPNER said the constancy of the sensibility of d'Arsonval galvanometers was a measure of the constancy of magnets having nearly closed circuits. Such instruments, in use at the Central Institution, had remained unchanged for several years.

Prof. S. P. THOMPSON admired Mr. Hibbert's instrument, and thought it would be very useful in laboratories. Standard cells, he said, were not always reliable, and condensers were the most unsatisfactory of electrical standards. On the subject of permanency of magnets, he said that Strouhal and Barus found that magnets with nearly closed circuits were most constant, and that, to give the best results, the hardness of the steel should be less the more closed the circuit. Mr. Hookham had also found that by using a nearly closed circuit and reducing the strong magnetisation by about 10 per cent, great constancy could be obtained.

Some years ago, he (Dr. Thompson) had tried the effect of ill-treatment on magnets, and observed that touching or hitting a magnet with non-magnetic material had little effect, whilst similar treatment with iron or magnets affected them considerably. Suddenly removing the keeper of a magnet tended to increase the magnetism, whilst putting a keeper on suddenly had the reverse effect. Strouhal and Barus had also investigated the temperature coeffcient of magnets, and found that this might be reduced by subjecting the magnet to rapid changes of temperature after the first magnetism, and then remagnetising.

Mr. W. WATSON enquired what was the percentage fall in strength of Mr. Hibbert's magnets. The bars used in magnetic surveys had been tested frequently, and they lost about o'5 per cent in six months.

The PRESIDENT asked what was the temperature coeffcients of the magnets described in the paper? Mr. Evershed, he said, thought it was between o'o per cent and 0.05 per cent for ordinary magnets. He thought the instrument shown by Mr. Hibbert would be of immense value if the magnet was really permanent. By it ballistic galvanometers could be readily calibrated, and, when combined with a resistance bore, it could also be used as a standard for current; for since the constant of a ballistic galvanometer for quantity can be determined from its constant for current, if the periodic time be known, conversely, that for current can be found from the constant for quantity. In some instances this would be of great use. Speaking of the temperature coefficient of condensers, he said that, in some cases, the specific inductive capacity of dielectrics diminished with rise of temperature, whilst in others it increased.

Mr. HIBBERT, in reply, said he found the temperature coefficient of his magnets to be roughly about o'03 per cent, but he had not investigated the matter very carefully. In making his measurements no correction had been made for the variation of capacity of his condenser with temperature.

Mr. WALTER BAILY, M.A., took the Chair, and The PRESIDENT communicated "A Note on Rotatory Currents.

The subject, he said, was probably familiar to most persons present, for it had been frequently referred to in the scientific papers. Alternate currents could be obtained fron an ordinary direct current dynamo by making contact with two points in the armature, say by connecting these points to insulated rings on the shaft, and using extra brushes. A direct current motor, similarly treated, trans. forms direct currents into alternating currents or into mechanical power. If two pairs of points in the armature be selected, situated at opposite ends of two mutually perpendicular diameters, then two alternating currents, differing in phase by 90°, can be obtained; and by choosing suitable points in the armature, two, three, or more currents, differing in phase by any desired angles, can be produced. In ordinary motors, the connections for doing this would be troublesome, but the Ayrton and Perry form, which has a stationary armature, lends itself readily to this purpose, for contact can be made with any part of the armature with great facility. A motor of this kind was exhibited, in which contact was made with four equidistant points on the armature. On connecting opposite points through fine platinum wires, and running the motor slowly, the wires glowed alternately, one being bright whilst the other was dark, and vice versa, thus demonstrating the existence of two currents in quadrature. When the four points on the armature were joined to the four corners of a square of platinum wire, the wires became incandescent in succession, the glow appearing to travel round the square, and suggesting the idea of rotatory currents. Tesla alternating current motor was also driven by two currents differing in phase by 90° obtained from the armature of the Ayrton and Perry direct current motor above mentioned.

A

The expression for shows that the phase of the current in circuit A, is independent of the resistance 1. On the other hand varying r2 alters . It was also pointed out that tan (+4) is generally greater than tan a.

Prof. J. PERRY, F.R.S., read a paper on "Struts and Tie-rods Laterally Loaded." He pointed out that in the case of struts a slight want of straightness may considerably reduce the breaking load. Even if a strut be originally straight and the thrust properly distributed, its weight usually produces lateral loading and consequent bending. Similarly, centrifugal force produces lateral loading in connecting rods. For some years the author has given his students practical examples of struts and tie rods to work out, taking into account the effect of lateral loads. The chief results obtained, together with a general treatment of the whole subject, are embodied in the paper. Where the curves of bending moment and the deflections due to lateral loading can be easily developed by Fourier's Series, solutions can readily be found. Simple cases of uniformly loaded struts and tie bars have been fully worked out, and also the case of locomotive coupling rods. In one problem on the latter subject a rectangular cross section was chosen and the proportions of depth to breadth determined so as to make the rod equally strong in the two directions when running at various given speeds. With cranks 12 inches long, the results show that at a speed of 390 revolutions per minute the ratio of depth to breadth must be infinite so as to give equal strength, so great is the influence of the lateral loading due to centrifugal force when combined with the thrust, Horizontal tie-rods loaded by their own weight have been investigated, and the tensions required to neutralise compression due to bending determined. A steel bar was used as a strut, with a thrust of 1500 lbs. The maximum stress due to bending by its own weight alone was 816, and on applying the thrust the maximum stress was raised to 23,190, or about twenty-six times that due to lateral loading alone. More complex cases have also been treated, the results of which are given in the paper.

OBITUARY.

PROFESSOR STAS.

THE chemical world-indeed, we may say Science in general has just experienced a severe blow by the death of Professor Stas, of Brussels. Although, perhaps, little known to the outside public, the deceased savant has earned undying honour and gratitude by his prolonged and splendid researches on the atomic weights of many of the elements. His laborious accuracy, his acuteness

NEWS

in detecting possible sources of error, his fertility of resource in their elimination, and his conscientious determination to be satisfied with nothing less than unquestionable truth, will remain a model for all generations. We purpose giving particulars next week.

NOTICES OF BOOKS.

An Introduction to Chemical Theory. By ALEXANDER SCOTT, M.A., D.Sc. London and Edinburgh : A. and C. Black.

OF elementary text-books, which are occupied almost exclusively by the enumeration of facts and the description of experiments, we have had of late years a plethora. But works which aim at bringing the results of observations and experiment into harmonious connection have been scarce indeed. Now, when even zoology and botany have left behind them the mere descriptive stage, it is surely an opprobium for chemistry to linger in the rear.

The author notices with due regret that "those who have had to teach the elements of both chemistry and physics to the same classes of students usually find the progress of their pupils more rapid and satis actory, for a time, at least, in the latter subject."

This difference he ascribes, with great probability, to the smaller number of facts, apparently disconnected, which are required in physics before they can be grouped together and built up deductively." His undertaking, therefore, is most commendable, and we believe that it will contribute to a wider and clearer understanding of chemical theory. Dr. Scott lays no claim, as regards the work before us, to either originality or completeness; but matter which is scattered through the scientific journals and the transactions of the learned Societies requires to be collected and often explained before it can be available

to the student.

In accordance with the destination of his book, the author confines himself within the boundaries of what may be regarded as fairly established, leaving speculative subjects for the study of another class of readers.

He arranges his subject-matter under the headings of The Constitution of Matter; Atomic Weights; Molecular Weights; Classification of Elements; Classification of Compounds, Acids, Bases, Salts, and Carbon Compounds; Abnormal Vapour-densities and Dissociation ; Physical Properties of Compounds; Thermo-chemistry; Chemical Change; Solution and Electrolysis.

Prout's

Dr. Scott recognises that the laws of Boyle, Charles, and Avogadro are true only to a limited extent. law is discussed at some length, mainly in order to throw light upon the determination of atomic weights. He considers that much may be said for this supposed law, if for the atomic weight of hydrogen we substitute part of the atomic weight of oxygen.

In considering the classification of the elements, he accepts the Newlands-Mendeleeff system, with certain reservations.

As regards formulæ and symbols, the author protests against their common use as a kind of shorthand to express indefinite quantities of the substances in question. This practice is very common in the abridged text-books now so abundant. He insists that each term of an equa. tion should refer to a definite number of molecules, and not of atoms as such.

Our ordinary notation, the author thinks, is sufficient for inorganic chemistry; but for organic compounds the case is different. Here, however, occur two questions:Does not the chemistry of silicon closely approximate to that of carbon? and, again, Can we suppose the atoms in each individual molecule arranged on a plane? Into this latter question the author enters to some extent when expounding the researches of Le Bel and Van 't Hoff.

In treating of solution, Dr. Scott decidedly leans to the

NOTE. All degrees of temperature are Centigrade unless otherwise expressed

Comptes Rendus Hebdomadaires des Séances de l'Académie des Sciences. Vol. cxiii., No. 22, November 30, 1891. On Boron Phosphides.-A. Besson.-The author mentions a curious reaction of boron phosphide. The product of its limited oxidation by means of dilute nitric acid, on evaporation to dryness, presents the aspect of nacreous laminæ, which, if dissolved in luke warm water, give with ammonia in excess a white gelatinous precipiThere appears to be formed under these conditions a peculiar phosphoboric acid, the ammoniacal salt of which is sparingly soluble.

tate.

Bromo-derivatives of Methyl Chloride.-A. Besson. The monobromo-compound, CH2BrCl, is a colourless liquid, which distils without decomposition at 68° with an odour resembling that of chloroform. It does not solidify at -55°. It becomes slowly coloured on exposure to air and light, with loss of bromine. Its specific gravity at +15° is 190; its vapour density taken with Hoffmann's apparatus in watery vapour is 4'72; theory demands 4'50. The corresponding bibromo compound, CHBr2Cl, has been previously obtained by Jacobsen and Neumeister in the action of potassa on a chlorobromic aldehyd. These chemists ascribe to the compound a boiling-point between 123-125°. This boiling-point must be lowered by several degrees so as to lie between 117-119°. It solidifies at 32°, its vapour density (taken in the vapour of aniline) is 7°18, the theoretical value being 7'25.

A Modification of M. Berthelot's Calorimetric Bomb, and on the Industrial Determination of the Calorific Power of Combustibles.-Pierre Mahler.This memoir does not admit of useful abstraction.

The Fixation of Free Nitrogen by Plants.-Th. Schloesing Fils and Em. Laurent.-The author's results obtain gaseous nitrogen from the atmosphere. Under are that there are inferior green plants which are able to the conditions of their experiments bare soils, oats, mustard, cress, asparagus, have not fixed free nitrogen in measurable quantities. It has been once more verified large quantities of atmospheric nitrogen. that under the same conditions peas are capable of fixing

Ammonia in the Atmosphere and in the Rains of a Tropical Region.-V. Marcano and A. Muntz.-The air in the tropical region studied (near Caracas, in Venezuela), is, contrary to expectation, less rich in gaseous ammonia than that of temperate climates. There is an abundant formation of nitric acid in the air of tropical countries, so that the ammonia combining with this acid ceases to be gaseous and forms crystalline dust.

Influence of the Solar Rays upon the Ferments met with on the Surface of Grapes.-V. Martinand.It is ound that grapes growing on the lower shoots of the vines yield a very large quantity of Saccharomyces, among which S. apiculatus predominates. On grapes from the middle height and from the summit of the vines, the ferments are very few in number. S. ellipsoideus, which is the most useful yeast for the fermentation of wine, is the more plentiful the less in.ense are the solar rays.