The brightest of the clusters are, of course, the nearest ones-w Centauri and Tucanae. They are easy naked eye objects and are about ten magnitudes brighter photographically than the faintest. These two leaders are far south; in fact, eighty per cent. of the globular clusters are in the southern hemisphere, as are most good things (astronomically speaking). This simply means that the solar system is inconveniently located pretty far to the north of the center of the Galaxy.

Iron and stone meteorites. There are on record fifteen times as many falls of meteoric stones as of meteoric irons or stony irons, according to a recent compilation by Miss Mussells, but the total weight of the irons is 2.5 times the total weight of the stones. This ratio of 2.5 may be taken as a measure, or at least as an intimation, of the relative frequency of iron and silicates in inter-planetary and interstellar space; it can be compared with the ratio of iron to stone in the Earth, which Dr. Fisher estimates at 4.5. Only the actual "falls" of meteorites are used in determining the ratio, for, once on the Earth's surface, irons persist but stones rapidly disintegrate, making estimates based on "finds" of meteorites of no significance for this computation."

The total number of recorded falls during the last four centuries, up to the year 1923, is 438. The total number of finds, where the actual fall was not observed, is 411. The total weight of all known falls and finds is some two hundred and fifty tons, a ridiculously small figure compared with the mass of the Earth, which is about 6 x 1021 tons.

New class O stars. There is a particular interest and usefulness in finding and investigating the superlatives the hottest or smallest or most distant or most something-or-other. It gives us an idea of the extremes of nature and of the scope of our problems. Likewise we seek to add new members to any rare or unusual class of objects in order to provide sounder data for statistical examinations.

The Class O spectrum appears to be a peculiar stage in the life history of very massive stars. Emission and absorption lines are there variously presented, and conditions of high excitation are indieated. The O's have been placed at the top of the temperature scale-perhaps erroneously-and the spectroscopists are not yet in altogether happy accord in matters relating to their classification and to their joining up with the B-A-F, etc., series of more normal and numerously represented spectral classes. We shall have more luck with more data in hand.

To the 140 already known, Miss Payne, in a current publication, adds ten new stars of Class O, two of which are naked eye objects. All are in the

southern hemisphere, and of course they lie close along the galactic circle, as do practically all stars of high intrinsic luminosity. These new members of the class have already brought up the question as to the proper limits between the late O stars and the early B's.

Beta Doradus. The discovery of a naked eye variable star with a large amplitude of light variation is certainly unexpected at the present time, for the sky has been studied thoroughly for decades by variable star observers and has been photographed hundreds of times. The far southern star ẞ Doradus, however, not only varies in a ten-day period throughout more than a magnitude, but its variation is of the Cepheid type, for which the light is never constant. A fourth magnitude Cepheid, such as this, pheids and also because these variables in general is important because of the scarcity of bright Ceare the most useful stars in the sky. They afford a powerful means of measuring great distances, and since their variations are not restricted to light alone, but affect color, velocity, spectrum and size, they bear on evolutionary problems. The Cepheids are exceedingly large-supergiants, we call them. And this one is a thousand times as bright as the Sun.

The discovery and measurement by Shapley and Miss Walton of the variability of ẞ Doradus on Harvard plates was brought about by reports on spectroscopic researches at the Lick Observatory, most recently by Miss Applegate.

The Lick Observatory, using plates made in Chile, showed that in spectroscopic behavior the star is allied to the Cepheids; the Harvard observers, using plates made in Peru, showed that the star is itself a Cepheid. The photographs that best show the variability are a special series of short exposures on the near-by Nova Pictoris. The brightness of ẞ Doradus had helped to conceal its variation heretofore, for on the usual photographs its images are overdone and so smeared that magnitude estimates are unreliable.

Now that the variability has been found and its period is known, we turn to the old star catalogues and find that the early visual observations had hidden in them the proof of the variability. The conservative astronomers of the past, however, had taken the blame on themselves for the large deviations shown by their measures, instead of mistrusting the

star.

Only two or three Cepheids are brighter than ẞ Doradus. The type star of the class, & Cephei, is one; and the North Pole star is another, but its variation is less than a tenth the amplitude of that of our new southern Cepheid.

Eclipse of the Moon. The lunar eclipse of June 15

was probably the best observed eclipse of the Moon in the history of science. It also brought about the highest degree yet attained in the cooperation of state, church, commerce and science in a single scientific problem. The Canadian and the United States Weather Services, the United States Army Signal Corps, the Roman Catholic Eskimo Mission, the fur traders and trappers along the Arctic Circle, the Royal Canadian Mounted Police, the powerful radio broadcasting stations of the Westinghouse Company, the astronomers of the observatories in western and southwestern America, the newspapers and Science Service, and the amateur astronomical observers over a large part of the United States were all involved in various phases of the meteorological and astronomical observations of this eclipse.

It will be many months before all reports come out of Alaska and from the Canadian Arctic giving information concerning the character of the atmosphere where the grazing solar rays passed and were refracted into the Earth's shadow cone to illuminate faintly, and discolor, the eclipsed Moon.

The general plans for the eclipse observations were developed by Dr. W. J. Fisher, of the Harvard staff, who has specialized in phenomena associated with lunar eclipses; the Canadian work was organized through the efficient cooperation of Dr. R. M. Stewart, director of the Dominion Observatory at Ottawa.

Nova in Magellanic Cloud. A nova in one of the Clouds of Magellan is of more than passing interest. In the first place, such an object has not heretofore been found in either of the Magellanic Clouds, notwithstanding the presence there of practically all other known types of high luminosity stars. In the second place, the distance of the Clouds are known, and therefore the actual luminosities of novae in such places can be computed, which is far from being the case for the nearer novae in our own Milky Way.

While comparing two photographs of the Large Cloud, in a study of stellar motions, Dr. Luyten recently noticed on a plate taken last September a star that was wholly absent from all of the many earlier plates. A search showed its images on eight other plates taken between September 28 and November 6. Just after this last plate was taken the Arequipa telescopes were dismantled for transfer to the new station in South Africa, and the further behavior of the star is unknown.

A cablegram to the South African observatories has resulted in some special photographs being made at the Union Observatory, Johannesburg, but the nova has apparently already disappeared. At maximum brightness the star was difficult enough, being photographically of the twelfth magnitude. But actually, if it is a member of the Magellanic Cloud,

it was, when brightest, some ten thousand times as bright as the Sun; the distance accounts for the apparent faintness. And while we are speaking of actualities we should add that it was only the terrestrial recording of a nova that occurred last September-the actual disaster happened nearly a thousand centuries ago.

HARLOW SHAPLEY

SPECIAL ARTICLES

FURTHER STUDIES ON THE ANTIRACHITIC ACTIVATION OF SUBSTANCES BY CATHODE RAYS1

In a previous preliminary report it was shown that, with the exposures used, high-voltage cathode rays applied directly were not effective in healing rickets in rats. On the other hand, it was shown that cholesterol could be endowed with antirachitic potency by exposure to the cathode rays. In our earlier attempt cholesterol was exposed to the cathode rays in a rather thick film and had to be added to the diet in amounts of 0.2 per cent. or more to bring about within two weeks complete healing of rickets in rats, which were rendered antirachitic by the Steenbock rachitic diet No. 2965.*

We have since found that, with the film of the substance about a millimeter or less, just as active products are formed by exposure to cathode rays as with ultraviolet irradiation and moreover the time interval is much shorter. In Tables I and II are summarized some of the recent experiments carried out. The cathode ray exposure was in all cases at a distance of 1 inch from the window of the tube and a current of 1 milliampere and 200,000 volts used. The substances were exposed in air and in one instance in an atmosphere of nitrogen. The cholesterol used in these experiments was a commercially pure product. In the experiments with cholesterol purified by the dibromide method, it was first brominated, then debrominated. brominated a second time, again debrominated and finally recrystallized three times from hot alcohol.

From an examination of Table I it is seen that cholesterol exposed to cathode rays for 30 seconds is ef

1 From the Laboratory of Biological Chemistry, Albany Medical College, and Research Laboratory, General Electric Company, Schenectady, N. Y. The writer is greatly indebted to Dr. W. D. Coolidge for his valuable advice and assistance, and for the technical assistance of F. S Randles and H. E. Tanis, Jr.

2 Knudson, Arthur, and Coolidge, W. D., Proc. Soc. Exp. Biol. and Med., 1927, XXIV, 363.

3 Coolidge, W. D., Jour. Franklin Institute, 1926, ecii. 693.

4 Steenbock, H., and Black, A., Jour. Biol. Chem., 1925, lxiv, 263.

2 mg. daily

Moderate

daily No healing

In Table II, it is seen that yeast, starch and cottonseed oil can also be activated by exposure to cathode rays. The experiments with yeast are extremely interesting. Yeast which contains ergosterol, considered to be the provitamin of vitamin D, is rendered very potent by exposure to cathode rays. With an exposure of 30 seconds, a dose of 1 mg daily brings about advanced healing of rickets within two weeks. Compared with a good grade of cod liver oil this is at least 10 to 20 times more potent. It is also interesting to note that a yeast which had been extracted previously for one week with ether in a continuous Soxhlet type of apparatus was still rendered very potent, indicating that the substance which is

* The line test showed beginning healing and the cornstarch in this case was exposed in rather thicker film than usual.

7 Rosenheim, O., and Webster, T. A., Biochem. Jour., 1927, XXI, 389.

In

The problem we are working with is a serious infectious leaf and top disease of the cane, called eye spot, caused by the fungus Helminthosporium sacchari Butler. Our first attempts against this disease were ineffectual, not from any difficulty in placing the dust on the cane but because the fungicidal dusts used caused little or no reduction of the disease. field-plot tests with adequate replications Bordeaux dust, and other copper mixtures, organic-mercury mixtures, sulphur mixtures, chlorine mixtures and coal-tar disinfectants gave little or no control of the disease. The best result in these earlier attempts was in plots treated with ordinary dusting sulphur in which we obtained 27 per cent. less infection than in alternating untreated plots which served as controls. This reduction of the disease was not sufficient to recommend for plantation practice but was sufficient to encourage us to seek further for more effective compounds.

Previous research by Young1 showed that the fungicidal action of sulphur was due to the formation of pentathionic acid formed by oxidation in the air. Young went further with this and secured greater fungicidal action by using more finely divided sulphur which would adhere to foliage better and oxidize more readily. Young's conclusions have been questioned in England, however.

The slight reduction of our disease with dusting sulphur nevertheless led us to follow up Young's work; instead of depending upon the oxidizing effect of the air, such oxidizing agents as nitric acid, one fourth of one per cent., and pulverized potassium permanganate, 1 per cent., were added to the sul

1 Young, Harry Curtis. "The Toxic Property of Sulphur." Annals of the Missouri Botanical Garden, Vol. IX, p. 403, 1922.

2 The Association of Economic Botanists. Discussion on "The Fungicidal Action of Sulphur." Ordinary Meeting, Oct. 20, 1925; The Annals of Applied Biology, Vol. 13, No. 2, p. 308, May, 1926.

phur. Seven plots of cane treated with the latter mixture have shown a reduction of 89.9 per cent. of infections as compared with seven undusted plots as controls; at the same time finely divided sulphur on seven plots has reduced the number of infections but 9 per cent. Sulphur plus one fourth of one per cent. nitric acid in eight similar plots reduced the disease 61 per cent. We have since increased the effectiveness of the oxidized sulphur even further by increasing the concentration of potassium permanganate to 5 per cent. No burning of sugar-cane foliage resulted, even when the concentration of potassium permanganate was increased to 10 per cent.

Potassium permanganate in a non-sulphur carrier such as kaolin has not reduced the disease as compared to untreated cane in control plots, indicating that the fungicidal effect is not due to the direct effect of the permanganate as a disinfectant but to its oxidizing effect on the sulphur.

It is possible to get quantitative data on the results of our treatments by marking an equal number of cane stalks in each plot and having counts of infections per leaf made at two-week intervals. The figures given above are from 140 leaf counts per treatment.

At the same time that our oxidized sulphur preparations gave a good control of the disease, we received a stimulation of growth, apparently independent of the disease-control, as shown by growth measurements of 70 cane stalks per treatment. The total increased growth was 8.8 inches per stalk, which would mean from two thirds to three fourths ton of sugar more per acre, a very profitable increase in yield.

It is our impression that dust fungicides against plant diseases have generally been less effective than liquid fungicides and that some of the sales of fungicide dust mixtures have been made by high-pressure salesmanship rather than on the basis of proven results in the field. We believe that our oxidizedsulphur mixtures will place fungicidal dusts more nearly on a competitive basis with liquid sprays, with the added advantage of greater economy of application. Care must be taken to keep these mixtures dry and away from fire, but millions of people have learned that gasoline can be used, with a few ordinary precautions.

Concerning mixtures of oxidizing agents in sulphur as fungicides, a patent has been applied for and if any royalties do occur they will be applied for the furtherance of research.

EXPERIMENT STATION OF THE HAWAIIAN SUGAR PLANTERS'

ASSSOCIATION

H. ATHERTON LEE J. P. MARTIN