61.6 to 62.2 per cent. of oxalate, (COO). This suggests the presence of a metal of higher atomic weight than calcium.

The filtrate from the calcium oxalate was evaporated to dryness, and ignited, leaving a very light brown residue, readily soluble in hot dilute hydrochloric acid.

When this solution was concentrated, short feathery needles first formed.

The hydrate of Dvi-manganese was precipitated from this solution by means of ammonium hydroxide. This gave a white flocculent precipitate which gradually darkened. The warm filtrate slowly deposited more of the brown precipitate.

This was all filtered off and the filtrate kept for further examination,

Properties of the Oxide. The precipitate of "dvi-manganese hydroxide "was soluble in hot dilute hydrochloric acid, but was much more easily dissolved by sulphurous acid.

Caustic alkalis in solution did not dissolve it. It also remained undissolved when sodium peroxide was added, although When it turned a dark chocolate colour. treated with sodium hypochlorite, solutions of dvi-manganese salts again gave a light which precipitate hydroxide darkened.

rapidly

or

When fused with caustic alkalis sodium peroxide it gave a bluish-green product somewhat resembling the colour of nickel compounds.

This was readily soluble in water, giving a deep green solution, which possessed oxidising properties. Thus it was decolorised by solutions of ferrous sulphate, oxalic acid, sulphur dioxide, and by carbon dioxide. Any manganate present would have been converted to permanganate by carbon dioxide and this was not observed.

The solution of potassium dvi-manganate was, however, unstable, and in the course of a day had become almost colourless, having deposited the hydrated oxide of the element.

Reactions of Dvi-manganese Salts.Like manganese dioxide, the dark brown oxide of dvi-manganese was readily soluble in a warm dilute hydrochloric acid and in sulphurous acid.

The solution gave purplish and greybrown precipitates with potassium ferroand ferri-cyanides, but no precipitate was obtained with sodium phosphate or monium oxalate.

am

Reference in the introduction has been made to our prior work, so it will only be necessary here to deal critically with the data thus far obtained by means of the X-rays.

Referring to Table II., this is inherently explanatory with the exception of one or two features. The values given in Column 3 are those in Siegbahn's book with the ex ception of those calculated for elements 75, 85. 87, 93, and the blank spaces where the radio-atoms fall. In Column 4 it will be seen that the differences are graded so as to fit into the gaps. Keeping to the difference value, 0.379, throughout a considerable range led to writing the values in Column 3 for thorium low and bismuth high, the Siegbahn values being respectively 30.916 and 28.259. This slight irregularity for the sake of uniformity does not appreciably affect the newly-calculated values.

In order to arrive at the Lẞ, values in Column 8, a curve was drawn as shown by Fig. 2, which is also self-explanatory. The broken line shows an alternative curve, the choice between the two being arbitrary.

Beginning at element of atomic number 75, the La, calculated wave-length agrees closely with one given by Berg and Tacke (see previous citation, Part I.), viz.,

340

made from pyrolusite, we got sufficient zinc present to produce strong Ka and Kẞ lines. Messrs. Hilger checked the corresponding L lines from element 75, along with those from copper against the zinc lines, and reported in favour of line A1.430 being that To of element 75. make sure that the copper anti-cathode was not emitting any other lines than those characteristic of copper, considering our long exposures, a 6-hour run was made with nothing on the face of the anti-cathode.

This is import

ant, because we feared that the La, line of mercury might account for the Lẞ, line we obtained as from the element 75 which averaged 1.2325, the HgLa, line being 1.2385. Our calculated value for the 75Lẞ, line was 1,2358 as against the calculated value given by Berg and Tacke, viz., 1.2352.

We have not ignored the possibility of a mercury Lo, line appearing, for the powdered material on the anti-cathode does not the always adhere sufficiently close to copper surface as to insure being kept down in temperature, though the anti-cathode is water-cooled. It will be remembered Aston (Phil. Mag., May, 1923), found that mercury was completely eliminated from the discharge by the presence of cadmium, and it did not reappear so long as the cadthe mium mirror [formed] remained on Similarly, walls of the discharge tube." S. Smith (Proc. Nat. Acad. of Sciences, U.S.A., 1924, X., 4), found that "both lead and aluminium formed bright mirrors in the nereby portions of the tube [used for exploding electrically wires], while tungsten produced a band of dark stain on the glass. When the pre-explosion vacuum of the wires produced was good, none This is sufficient gas to pass a discharge. of especially interesting in the this metal is ordinarily aluminium, as thought to contain much occluded Possibly the gas is given off and then immediately reabsorbed by the thin film of aluminium on the walls of the tube." The

case

gas.

in our case mercury

as

absorption of gas, in vapour, in and by the presence of certain metals that partially vapourise on the anticathode, might afford a means of getting the La, line of mercury, as mercury is employed to exhaust the tube used, but the L spectrum would then be excited, we should expect the B, line to show as it has a strength of 8. No such line apeared when those of 75 were present. Messrs. Hilger regarded the a and ẞ lines clearly registered on several films as those of element 75, as the foregoing reasoning indicates.

Considering now element 85, the evidence in this case is fairly good, but the lines were not as well-defined as one would wish.

As regards element 87, there is an appreciable difference between our calculated La, line, 1.0276, and the observed line, 1.032 (faint, but clear, and the only one on the film except those of 75 and copper), the difference being 0.0044, which is about our experimental error. Of course, the calculation of the La, line may be a little low, as it occurs in the middle of the blank region. We have connected the observed line with the mean between the La, and La, lines as the line showed quite a perceptible width, but it is to be expected that the former will have a strength of 10 as against 3 for the latter, so that, in measuring, the middle of the La, line would be taken. As explained before, the Br-Ag absorption region seemed to block out the Lẞ, line if present, moreover, its strength would be slightly less than the La, line.

The last element of all, of atomic number 93, was discussed in our first communication (Part I., this Journal, October 30, 1925), the possibility of this element existing having stimulated our research from the start. There is some evidence of its existence, but to be critical, the line 0.693 comes near to the limit of the region explored by the setting (oscillation) of the rocksalt crystal of the spectograph, and edge effects are here. At any rate, this may acount for some of the extreme lines given at the foot of Columns 12 and 13. However this may be, the values recorded as lines by Messrs. Hilger are very close to those calculated, that is to say, they are within our experimental error.

66

possible

Films 1 to 3 are taken with the radiations of the first sample, which was made from manganese sulphate as described in

Part I. Films 4 and 5 are from a later sample made from pyrolusite.

Since the foregoing was written, we have a film from the sample made from pyrolusite, above referred to, which gives distinctly but faintly the lines 0.888 and 0.897 as measured by Messrs. Hilger. One line seems to correspond with the La, line of element of atomic number 93. Further X-ray spectroscopic work is in pro

gress.

All values are, where the context admits, wave-length in Angström units, as before.

STERILISATION OF TOWN'S WATER.

some

THE USE OF CHLORINE GAS.

(FROM A CORRESPONDENT.)

extent.

we

At the 14th Annual Conference of the British Waterworks Association, held at the Guildhall, London, on the 1 July last, a particularly interesting paper was read by Sir Alexander Houston, the Director of Water Examination, Metropolitan Water Board, on the subject of the purification of London's water supply. This dealt with many different aspects of the complicated question of supplying over 7,000,000 people with household water, but one of the most interesting was that of sterilisation by means of a measured trace of chlorine gas, which reduces very considerably the reservoir accommodation necessary since the natural slow process of sterilisation by means of storage is thereby eliminated to Thus, for example, learn from a previous report of the Metropolitan Water Board, about 76,000,000 gallons of London water is treated daily with chlorine, which, at the consumption of 38 gallons per head, corresponds to the requirements of 2,000,000 people, that is nearly one-third of the population of Greater London. The results have been highly satisfactory, although the magnitude of the problem of London's water supply is well illustrated by the new reservoir at Littleton, near Staines, which is one of the largest in the world. The capacity of 6,750,000,000 gallons is enough for a month's supply, the surface of the reservoir being 725 acres, the distance round about 4 miles, and the cost £2,050,000, the total reservoir accommodation of London being now about 14,000 million gallons, although it would have to

be considered greater if it was not for the chlorine treatment.

There is no question that the best way to use chlorine is in the form of pure liquefied chlorine gas, about 1 part per 4,000,000 parts being the usual dose, although this varies somewhat depending on the iron and the dissolved organic matter content, which also absorb chlorine. The use of the gas instead of hypochlorites was first suggested in 1903 by Lieutenant Nessfield, of the Indian Medical Service, and applied in 1910 by Major Dornell, M.D., of the United States Army Medical Corps, and it has the great advantage of eliminating the inherent objection to the use of bleaching powder or sodium hpyochlorite, that of unstability, especially in warm weather. As is, of course, well known to chemists, the percentage of active chlorine, instead of always remaining at about, say, 35% for solid bleaching powder, and 14-16% with sodium hypochlorite solution, is apt to diminish to a considerable extent, renderit almost impossible for the attendant to control the exact amount of chlorine being supplied to the water. It is for this reason that chlorinated water sometimes tastes and smells slightly, since in order to make certain of complete sterilisation, excess of the reduced hypochlorite is added. The use of pure liquefied chlorine gas in cylinders, however, gets over this difficulty, since it is always pure 100% chlorine, and remains unchanged under all climatic conditions, whilst also it has the further advantages of not adding any soluble salts to the water. Chlorine gas is used by means of the "Chloronome " apparatus of the Paterson Engineering Company, Ltd., of London; and other important points are easy adjustment and continuous automatic addition of the chlorine in the most accurate character, so that there is not the slightest taste or smell in the treated water, the cylinders of chlorine being conected to the "Chloronome apparatus which by means of suitable reducing, regulating, and control valves, passes continuously a small stream of the gas, instantly adjustable in amount by one stop valve. This gas is absorbed by a trickle of water in a small absorbing tower forming part of the apparatus, and the dilute solution of chlorine passes into the main water supply, mixing almost instantly, a very important feature in practice, since it ensures absolute equal chlorination of the whole bulk of the water, otherwise not easy to obtain.

THE VITAMIN CONTENT OF CORTEX LIMONIS B.P.

Anent the abstract of the work of Dr. Juritz on the Citrus family, published in the Chemical News on November 13, we have been favoured by the authors, Stanley G. Willimott, Ph.D., B.Sc., A.I.C., and Frank Wokes, B.Sc., A.I.C. (pharmaceutical chemist), with a copy of an interesting paper under the above title, read before the Science Section at the British Pharmaceutical Conference at Glasgow, on July 30, 1925, and reprinted from the Pharmaceutical Journal and Pharmacist of August 1, 1926. The paper deals fully with most phases of the subject, and will constitute a valuable addition as a work of reference on an important subject. We give attached two excerpts, one dealing with the historical aspect, and the other with the principal constituents.

warmer

Although lemons have been known from very early times, their use in medicine has arisen only quite recently. It seems probable that the genus Citrus originally came from Eastern Asia, and a number of species are now found inhabitating the valleys of the Himalayas. Strasburger (4th English edition, 1912, p. 604) states that all the important cultivated forms have been obtained from the Chinese. Citrus medica is the form which was known to the Greeks in the expeditions of Alexander the Great as the Median apple. Lemons were only known to the Romans at a late period in their history, and were at first used only to keep the moths from their garments. Pliny mentions their effectiveness as an antidote to poisons. The great obstacle to the wide use of lemons was the difficulty of preserving them for any length of time; also their juice was disagreeably acid and would not keep, and this was not overcome until sugar became available in large quantities in the seventeenth century. A syrup of lemon juice was then made in Italy, and first imported into England somewhere about 1630. Meanwhile, the value of lemon juice in preventing scurvy, a disease dreaded by all who had to take long voyages in sailing vessels, was gradually becoming recognised. Lind ("A Treatise on the Scurvy," London, 2nd edition, 1757) recounts the tragic history of four ships which sailed from England to Bombay in April, 1600, carrying 480 men to establish the East India Company. The Commodore on his own ship had arranged

for a regular issue of lemon juice, three tablespoonfuls daily to all hands, and at the end of the voyage his men were all in good health. On the other three ships, where no lemon juice had been issued, 105 men had died of scurvy, and the remainder were so weak that they could not unload their cargoes. Captain Cook, in his famous voyages in the following century, gave excellent examples of the prevention of scurvy by means of lemon juice, and it was probably this that led to its being made a regular issue in the Navy in 1804, after which scurvy, which had been responsible for thousands of cases annually, became The value of the comparatively rare. lemon juice was thought to be due to its acidity, and when it was later found that the juice of the West Indian lime (Citrus medica v. acida) was more acid than that of the lemon (Citrus medica v. Limonum), lime juice was gradually substituted. Thus lime juice has ursurped the reputation of lemon juice as an anti-scorbutic, although as a matter of fact lemon juice is more than twice as effective (see table drawn up by staff of Lister Institute and published in the Medical Research Committee's Report on th present state of knowledge concerning accessory food factors, p. 44).

Although the juice was thus the first part of the lemon to be used medicinally, the virtues of the peel were not altogether ignored. Thus, in Dr. Nicholas Lemery's

Course of Chymistry," 3rd edition, London, 1698, p. 506, are given directions for the preparation of a Magistral or Compounded balm-water.

PRINCIPAL CHEMICAL CONSTITUENTS.

In the oil are found the terpenes dlimonene a-terpinese, and a-phelland rene, the first being in largest amount, and the three totalling about 90 per cent. of the bulk of the oil; the aldehydes citral (minimum 4 per cent.) and citronellal; and the ester geranyl acetate, while linalyl acetate is said to occur in Palermo oil. The mucilage accompanying the oil requires further investigation.

Another important constituent is the glucoside hesperidin. This substance when hydrolysed yields hesperetin, rhamnose, and two molecules of glucose. It thus affords the plant a possible means of storing up the reserve energy represented by its reducing sugars, which, as has been shown above, it actually does in the un