the thirty-three years 1871 to 1903, based upon the observations for London (Brixton) published in the daily weather report of the Meteorological Office. The mean annual number of foggy days is 55, of which 45 occurred in the winter half of the year. Dividing the thirty-three years into three equal periods, the result is, for the first period, a mean of 55, for the second 69, for the third only 41. Since the year 1888 a steady and uninterrupted decrease is shown in the mean annual number of fogs. Among the principal agencies which may have conduced to this desirable result must be mentioned the efforts of the Coal Smoke Abatement Society and the London County Council, also the use of incandescent gas light and electricity; but, as pointed out by Captain A. Carpenter and Mr. C. Harding, the increase of wind in recent winters is probably chiefly responsible for the decrease of fog. As we have remarked before, the geographical situation of London is, from a purely meteorological point of view, eminently favourable to the development of fog, and the only permanent improvement we can hope for is an abatement of its more injurious effects caused by the imperfect consumption of coal and gas.

HAVING Occasion recently to devise a short-focus spectrograph, Prof. Wood, of the Johns Hopkins University, found it necessary to make a study of the distribution of light (monochromatic) in the different orders of a typical grating. His method, a beautifully simple one, is described and illustrated in No. 2, vol. xxi., of the Astrophysical Journal. The result showed that, in the typical grating experimented with, half the reflected light was concentrated in one spectrum, and as the grating reflected about 76 per cent. of the total incident light, this means that about one-third of this total was found in the one spectrum, which was one of the two first orders. It was also found that the ruling makes little or no difference to the total reflecting power of the speculum. Two flint prisms of 60° would give about the same average dispersion as that produced, and, according to Pickering's table in Kayser's "Handbuch," they would transmit a little more than twice the light reflected, in the first order of the grating used.

THE Psychological Bulletin, ii., 2, contains reports of the proceedings of the thirteenth annual meeting of the American Psychological Association and of the fourth annual meeting of the American Philosophical Association, which were both held at Philadelphia on December 28-30. Abstracts of the papers are given. Invitations on behalf of Harvard University to hold the next annual meeting in Cambridge, Mass., to signalise the opening of the Emerson Hall of Philosophy were accepted by both associations, and it is proposed that the Western Philosophical Association and the Southern Society for Philosophy and Psychology shall also meet at the same time and place.

A COLOURED plate of a new species belonging to a new genus of Hydrachnidæ is given in the Rendiconti of the Lombardy Institution, xxxviii., 3, in illustration of a note by

Mr. R. Monti on the new "find." This water mite was obtained in cold springs on the right bank of the Anza, near Ceppomorelli, and has been named Polyxo placophora. The same writer in another number of the same journal discusses the horizontal migrations of lacustrine plankton, and finds in mountain lakes that, in addition to the known vertical movements, there are well-marked diurnal migrations of the small crustacea to different parts of the lake depending on sunshine and shade.

IN the March number of the American Journal of Science Mr. Charles S. Hastings utilises some observations of the power of accommodation of the eye for light of different wave-lengths to make a complete determination of the optical constants of the eye for all conditions of accommodation and for all colours. The results are given in two tables, by the use of which all problems connected with the purely optical properties of the schematic eye may be solved.

a

IN the course of an investigation of radio-active muds which is published by Prof. G. Vicentini in the Atti of the Royal Venetian Institute (vol. Ixiv., ii., 535), the connection existing between the ionisation produced by the mud and the quantity of material used is experimentally ascertained. When the mud is spread uniformly over definite area, the intensity of the radiation increases as the thickness of the layer is increased, but a direct proportionality does not exist between them. After a certain point, moreover, the radio-activity is not increased by adding fresh material. Mr. H. S. Allen, in a paper read before the Royal Philosophical Society of Glasgow on January 25, also deals with radio-active water and mud, the material in this case being derived from the springs of Bath and Buxton. An interesting point which is established incidentally is that the fluorescence excited in a sensitive plate by the radium rays plays only a very minor part in the production by these rays of a photographic effect.

AN interesting investigation of the secondary radiation produced when the 6 and 7 rays of radium impinge on metallic plates is published by Prof. J. A. McClelland in the Transactions of the Royal Dublin Society (vol. viii., No. 14). It is shown that the secondary rays are not produced merely at the surface of the plate struck by the primary rays, but that they come from all parts of a layer of considerable depth. Apparently the less penetrating B rays are more efficient in producing a secondary radiation than the y or highly penetrating rays. The nature of the secondary radiation depends largely on the character of the metal employed; the greater the atomic weight of the latter the greater is the amount of the secondary radiation produced by it. Of all the substances experimented with, lead gives rise to the greatest effect, both as regards the quantity of the secondary radiation and its penetrating power. The secondary radiation consists, apparently entirely, of a species of B rays, that is, of negatively charged particles capable of deflection in a magnetic field. Perhaps the most important feature of the paper lies in its directing attention to the necessity of considering secondary radiations in all measurements of the absorptive power of substances with regard to the rays produced by radioactive bodies.

WE have received a copy of a memorandum on the construction and verification of a new copy of the imperial standard yard, by Mr. H. J. Chaney, superintendent of the Standards Department of the Board of Trade. Since the original standard yard of bronze was made some sixty

544

years ago, it has been found that bars which are constructed of copper alloys do not retain their original length with that degree of accuracy now demanded for scientific purposes. The new copy (I.P.) is made of an alloy containing 89.81 per cent. of platinum and 10-10 per cent. of iridium, such an alloy being little affected by changes of temperature and not at all by oxidation; as the alloy admits of a high specular polish, the fine lines marking the extremities of the yard can be traced directly on the bar without the intervention of gold plugs or pins as in the older type. Instead of using the old solid 1-inch section, for the purpose of lightness the so-called "Tresca" section has been adopted. The memorandum gives full details of the verification of the length and a description of the apparatus used, including the thermometers by which temperature was measured and a new microscopic "comparator" similar to that used at Paris by the Comité international des Poids et Mesures. This instrument has been purchased by the Board of Trade and mounted in a special chamber at Old Palace Yard, Westminster.

VESSELS of fused quartz can now be obtained commercially, and on account of the remarkable properties of this substance, a wide field of research at high temperatures would appear to be opened up by their use. In high temperature gas thermometry, for example, where glass is excluded on account of its comparatively low melting point, and platinum on account of its permeability to hydrogen, fused quartz promised to be an ideal envelope. Unfortunately, Villard has found that fused quartz is also permeable to hydrogen at high temperatures, well below its melting point, and Jacquerod and Perrot have proved that helium resembles hydrogen in this respect. In the current number of the Comptes rendus (March 27) M. Berthelot shows that the use of quartz vessels is still further limited, as both oxygen and nitrogen can penetrate into hermetically sealed quartz bulbs at 1300° C. Thus carbon, heated in sealed vacuous quartz tubes for half an hour at 1300° C., gave a mixture of nitrogen and carbon monoxide on cooling the tube and extracting the gases. Experiments were made on other substances, and all the facts pointed to the conclusion that at a high temperature fused silica behaves towards gases like an animal membrane, susceptible of endosmosis and exosmosis, the phenomenon depending partly on the thickness of the wall. It is clear, therefore, that before this substance can be used with confidence in high temperature work, a further study will have to be made of its defects in this direction.

THE Comptes rendus for March 27 contain an interesting paper on the cryoscopic behaviour of hydrocyanic acid, by M. Lespieau. According to the views of Nernst and Thomson on the relation between the dielectric capacity and the power of electrolytic dissociation, the fact that the dielectric constant of prussic acid is higher than that of water should give the acid a higher dissociating power. M. Lespieau has accordingly carried out a series of experiments on the lowering of the freezing point of this substance by the addition of alcohol, chloroform, benzene, water, trichloracetic acid, sulphuric acid, potassium iodide and nitrate, and has found that for the first six substances the cryoscopic constant is between 19 and 20, whilst for the two latter it is approximately double. Hence the two acids, which are strongly dissociated in water, are not sensibly dissociated in prussic acid solutions of the same strength, and this is in accord with the experiments of Kahlenberg, who found that these solutions were bad con

=

COMET 1905 a (GIACOBINI).-A second telegram from the Kiel Centralstelle announces that comet 1905 a was observed by Prof. Aitken at Lick on March 27. The position at March 27d. 7h. 57.1m. (Lick M.T.) was R.A. = 5h. 48m. 55s., dec. +12° 35' 43". Apparently, then, the northern declination is increasing, An error in the and not decreasing as previously stated. key by which the code telegrams are translated substituted declination for N.P.D., so that the daily movement in declination should be read as plus 1° 15'.

The following elements have been computed by Dr. E. Strömgren from observations made on March 26, 28, and 30, and are given in Circular No. 76 of the Kiel Centralstelle, together with a bi-daily ephemeris extending from March 30 to April 23

French

PHOTOGRAPHY OF THE CORONA WITHOUT A TOTAL ECLIPSE. to the note communicated -According to a Academy of Sciences, and in the opinion of M. J. Janssen, M. A. Hansky has succeeded in photographing the corona of the uneclipsed sun. The photographs were taken with a 12-inch telescope in the exceptionally transparent atmosphere which obtains at the observatory situated on the summit of Mont Blanc.

After a number of preliminary experiments on the selective absorption of screens dyed with various aniline colours, M. Hansky obtained a combination which absorbed all radiations more refrangible than 660 μμ, and, as the red radiations of the corona are very intense and do not suffer absorption or dispersion in passing through the terrestrial atmosphere, he used this screen in obtaining twelve negatives. The individual screens were prepared by soaking a fixed undeveloped Lumière film in each of the suitable dyes, and, between each exposure, they were re-arranged inter se so that no false effect due to any particular disposition of the grain might affect the resulting picture. The direct photospheric and chromospheric rays were prevented from reaching the plate by

66

the interposition of a blackened brass disc slightly larger than the solar disc.

The resulting negatives showed distinct halos around the disc, and, notwithstanding the fact that some time elapsed between the successive exposures, these halos exhibited the same form, thus testifying to their solar origin. Some of the negatives were photographically intensified by repeated copying, and reproductions of them were submitted to the academy. In presenting the communication M. Janssen-to whom M. Hansky acknowledges his obligations for assistance and advice-stated that "the photographs actually show the solar corona with an intensity and a perfection only known on the photographs obtained during total eclipses (Comptes rendus, No. 12).

SEARCH-EPHEMERIS FOR TEMPEL'S FIRST PERIODIC COMET (1867 II).-Although Tempel's first comet has not been seen during its last three perihelion passages, i.e. since 1879, M. A. Gautier, of the Geneva Observatory, thinks that the probability of its re-discovery this year is great enough to justify a careful search. For this reason he re-publishes, in No. 4008 of the Astronomische Nachrichten, the elements he prepared for the 1898 apparition, reduced to the mean equinox of 1905.0. As the probable time of perihelion is somewhat uncertain, he gives three ephemerides, extending from March 31 to July 13, in which this time is reckoned as May 2-5, April 20.5, and April 8.5 respectively, the mean date being the most probable. The declination varies from -16° to -31°, so that the more southerly observatories are more likely to be successful in the research.

RIGHT ASCENSIONS OF 2120 SOUTHERN STARS.-In an appendix to "Observations made at the Hong Kong Observatory during 1903," Prof. W. Doberck, the director, publishes the right ascensions of 2120 southern stars for the epoch 1900, as determined from observations made by Mr. J. I. Plummer and himself during the years 1898 to

1904.

The observations were made with a 3-inch Simms semiportable transit instrument, which, together with the method of reduction and the comparisons with other catalogues, is briefly discussed in the director's preface.

In the catalogue itself, the number of the star as given in Lacaille, or Stone, or both, the R.A., epoch and magnitude, the variation of the R.A. from Stone's corresponding value, the proper motion, and several other particulars are given for each star.

THE IRIS DIAPHRAGM IN ASTRONOMY.-In a communication to the French Academy of Sciences, M. Salet states that he has recently and usefully adapted the iris diaphragm to a telescope in which the magnification employed is 500. The diaphragm is placed very near to the plane of the micrometer wires in front of the field lens, and its raison d'être is to prevent the light from the sky, and from the illumination of the wires, from reaching the eye when feeble objects are being observed, the diaphragm being closed by an external cylinder when the object has been brought to the centre of the field. By reducing the extent of the micrometer wires, the diaphragm also reduces, or eliminates, the effect of astigmatism when observations of double stars are being made (Comptes rendus, No. 9).

CONSTANCY OF "SPARK " WAVE-LENGTHS.-A question which is of first importance to those observers engaged in stellar line-of-sight work, viz. that of the constancy of wave-lengths in spark spectra taken under various conditions of discharge, has recently been re-investigated by Mr. G. W. Middlekauff at the Johns Hopkins University. A detailed description of the apparatus and methods employed, together with the results obtained, appear in No. 2, vol. xxi., of the Astrophysical Journal.

Mr. Middlekauff used a concave Rowland grating of 20,000 lines to the inch and a focal length of 21.5 feet. The self-induction in the spark circuit could be varied from 0.00007 to 0.0012 of a henry, and the capacity from 0.0085 to 0-0739 of a microfarad, and the results obtained afford strong evidence that in the case of a spark discharge in air, at atmospheric pressure, no "shift in wave-length is produced by variations of self-induction or capacity

within the above limits. A further result obtained was that the analogous wave-lengths in the arc and the spark spectra of the same elements are not measurably different.

STATISTICS OF VARIATIONA

A PAPER consisting mainly of a large number of elaborate records bearing on the important subject of variation has recently been issued by the Washington Academy of Sciences. The data, which have been collected with much care and industry, cannot fail to be of high interest to all students of evolution. They afford an excellent example of the peculiar value of insect studies in reference to many difficult problems in biology-a point Prof. which has lately received fresh emphasis from Poulton's valedictory address as President of the Entomological Society of London.

The authors start with an "Introduction," in which they declare their "belief in the marked betterment and effectiveness of practically all variation study when pursued from the point of view of the biometrician "; adding, however, that" from the writers' point of view the study of Dealing with the special advantages presented by insect variation is a phase of biology, and not of mathematics." data in this inquiry, they assert that the phenomena of complete metamorphosis afford a ready means of distinguishing "variations which are strictly blastogenic from others which may be in large part acquired.' This, it may be remarked, is only true under certain limitations. It is not the case, for instance, as the authors appear to think, that the imaginal colour-patterns of lepidoptera are uninfluenced by the conditions obtaining during individual development.

Coming now to the main substance of the paper, we find a series of short articles or sections giving statistics of variation in some two dozen species of insects. Among the structures thus dealt with are the venation and costal wing hooks in bees and ants, the venation in gnats, the colour-patterns of sundry beetles, wasps and bugs, the eye-spots of certain butterflies, the tibial spines, tarsal and antennal segments, tactile hairs and elytral striæ of other insects of various orders. In the case of the hive bee it is incidentally shown that the parthenogenetically produced drones are as subject to variation in their wings as are the workers of biparental ancestry. The results are in many cases graphically summarised in the form of the frequency polygon; and the mode," "standard deviation,' "index of variability," and "coefficient of variation are duly reckoned and recorded in accordance with approved biometrical methods. It is interesting but not surprising to observe that the frequency curve is usually in fair correspondence with the law of error.

66

The paper ends with a section devoted to "general results. Here we think that too much is made of the difficulty of distinguishing between congenital variation and acquired modification. For practical purposes the distinction is usually obvious enough. A little later the authors observe, The most satisfactory answer to the question of the hereditary transmission of acquired characters will come as the result of a quantitative (statistical) study of variations known to be blastogenic compared with a similar study of variations known to be acquired, both studies to be made on complete series of individuals bred This under quantitatively determined life conditions." seems to us somewhat like using a steam-hammer to crack an egg. It is not astonishing to find that there is little or no evidence of differing selection-value in the variable number of spots on the elytra of a ladybird; but it hardly seems clear that the authors are justified in claiming this fact, together with mode an apparent change of between the years 1895 and 1901, as evidence in favour of "determinate variation." Before any such inference can properly be drawn, the question of possible correlation ought at least to be considered. The authors, however, arrive on the whole subject at the satisfactory conclusion that natural selection "is after all a logical necessity and undoubtedly an actual actively-regulative factor" in the formation of species. F. A. D.

1 "Studies of Variation in Insects." By Vernon L. Kellogg and Ruby G. Bell, of Leland Stanford Junior University. From the Proceedings of the Washington Academy of Sciences, vol. vi. (Washington, D.C., 1904 )

INTERRUPTERS FOR INDUCTION COILS. IT has been thought that an account of the more important forms of interrupter would not be unwelcome to readers of NATURE.

A rotating air-break interrupter is shown in Fig. 1. An accurately balanced brass fly-wheel, FW, driven by a small motor, is fitted with two insulating segments, Is, let into its periphery. Bearing on the fly-wheel are two copper gauze brushes, B, and B2; the circuit is interrupted as each brush slips over from the brass to the insulating

portion of the rim. It is evident that the arcing which occurs at the break necessitates the use of a fire-proof insulator. A small piece of slate (s in Fig. 1) is fitted immediately behind each brass segment, and this takes the spark; it is easily renewed, the remainder of each insulating segment consisting of vulcanised fibre.

So far as the writer is aware, this type of interrupter was first described by Wadsworth in 1894, and was used by Prof. Michelson in some Geissler-tube experiments (American Journal of Science, pp. 496-501, December, 1894). As might be expected, the suddenness of the break depends

any considerable wear are the slate distance-pieces; the rim of the fly-wheel may occasionally require truing-up. It is important to keep the edges of the brass contactsegments and the surfaces of the slate distance-pieces clean by the occasional application of fine sand-paper.

In Fig. 3 are shown the essential parts of the mechanism of a double-dipper interrupter. The double motordriven crank, c, carries two connecting-rods, CR, each of which is attached to a cross-head, c H. Each cross-head is guide-springs, Gs, and through the guide-block, GB. The fixed to the top of a stiff rod, R, which passes between the latter is supported by a strong bracket, B, screwed to the stand supporting the motor. Each reciprocating rod ends in an amalgamated copper wire, cw, which dips into the mercury. It will be readily seen that by the adoption of the two-crank arrangement the frequency is doubled for a given speed as compared with the single-crank interrupter; for while with the latter there is only a single break per revolution, the former gives two breaks per revolution, one of the contact-rods or dippers entering the mercury shortly after the other has left it. The mercury cup itself is made adjustable in a vertical direction, and is, as usual, immersed in alcohol.

66

The curve marked "double dipper" in Fig. 4 gives the results of a test with this form of interrupter. The frequency of interruption was 22. The results correspond fairly well with those plotted in Fig. 2 for the rotary air-break interrupter at a frequency of 40.

This type of interrupter is comparatively cheap and simple, and works very steadily. There is no complicated mechanism to get out of order, and only a small quantity of mercury is required (about 2lb.).

One of the most successful types of rotary interrupter is the mercury jet interrupter. Several varieties of this have been used. One of the best known is shown in Fig. 5. The vertical motor-driven shaft, S, carries a cylinder, c, the lower portion of which is cut up into a number of teeth, T. The shaft s is continued downwards, and passes through the mercury pump casing. The mercury pump is of very simple construction, and is shown in Fig. 5 (b). Inside a flat oval box, which forms the pump casing, are arranged two thick toothed wheels. One of these is mounted on the lower end of the shaft s, which carries the toothed cylinder, Fig. 5 (a), and drives the other. The wheels fit the inside of the casing very closely, and are arranged to rotate as indicated by the arrows in Fig. 5 (b). The mercury imprisoned

CW

FIG. 3

GS

between the teeth of the wheels and the casing is consequently carried round and forced through the nozzle. The issuing fine jet of mercury, MJ-Fig. 5 (a) is directed against the rotating teeth, the break taking place at the vertical edge of a tooth. The height of the nozzle N is adjustable, and by this means the magnitude of the current may be regulated, as by raising the nozzle the jet will be directed against a tooth for a longer period, and the current will attain a larger value before the break takes place. The entire mechanism of this interrupter is contained in a strong cylindrical glass vessel, the lower portion of which contains mercury, in which the pump is immersed, and with which the pump chamber freely com municates by means of a suction orifice, while above the mercury is the usual alcohol filling the bulk of the vessel. If in good working order, the mercury jet interrupter gives excellent results, as may be seen by referring to the curve marked "mercury jet" in Fig. 4, which corresponds to a frequency of interruption=40. A comparison of this curve with that given in Fig. 2 for the rotating air-break interrupter at once shows the superiority of the jet interrupter. The mercury jet interrupter is much more expensive and complicated than the " double-dipper" type, and requires a larger amount of mercury; but it yields somewhat better results.

interrupter is to be supplied. The interrupter is connected in series with the primary of the induction coil, and, if necessary, with an additional self-inductance. As soon as the circuit is closed, and provided the area of anode surface exposed to the electrolyte is not excessive, and the self

Another form of electrolytic interrupter, originally due to Caldwell, but subsequently improved and modified in various ways by others, is shown in Fig. 7. The terminal T2 is, as in the Wehnelt interrupter, connected to a lead plate. But instead of a platinum anode, a lead plate is also used for the other electrode. This second lead plate is surrounded by a glass tube, GT, which completely separates it from the remainder of the electrolyte except for a small perforation at the bottom of the tube, through which passes the pointed end, P, of a long glass rod, G, supported in a tubular rack rod, R, which may be raised or lowered by means of a pinion fitted with the milled head, м H. The area of communication between the electrolyte in the tube and that outside is controlled by raising or lowering the conical glass plug. Either electrode may be used indifferently as anode or kathode. The break takes place at the perforation of the glass tube.

In conclusion, thanks must be expressed to Mr. A. C. Cossor, of 54 Farringdon Road, E.C., who very kindly provided an induction coil and a number of interrupters required to carry out the tests recorded in this article.

UNIVERSITY AND EDUCATIONAL
INTELLIGENCE,

IT is stated that Sir William MacDonald, of Montreal, has decided to give 800,000l. toward the erection of a normal school at St. Anne de Bellevue, a few miles distant from Montreal, and the erection and endowment of an agricultural college at the same place.

THERE is no sign of diminution in the interest shown by public authorities and by private benefactors for higher education in the United States. We learn from Science that by the will of Mrs. Stanford about 400,000l. is bequeathed to Leland Stanford Junior University. The university also comes into possession of the house built by Senator Stanford at San Francisco and its contents, which are valued at more than 400,000l. The legislature of North Carolina has appropriated 10,000l. for the erection of a chemical laboratory at the University of North Carolina.

WE have received a copy of the prospectus of courses of instruction in poultry-keeping held at University College, Reading, and the college poultry farm at Theale. The farm, which is of about 40 acres, largely meadow land, is used also as an experimental station. The courses are of varying lengths and different degrees of difficulty to meet the requirements of all grades of students. The practical work is exhaustive, and due attention is given to kindred technical subjects such as carpentry. It appears that this branch of the work of the college has had an important influence on the development of scientific poultry-keeping in Berkshire and neighbouring counties.

A STRONG committee has been formed for the purpose of securing suitable conditions of work, and providing opportunities for development, of Bedford College for Women in London. An appeal to the public on behalf of the college has just been issued. The college, which is a school of the University of London, must before long come to an end unless it can obtain a large amount of public support. A freehold site and a new building are essential, and it is estimated that their cost may amount to 150,000l. Experience has shown that the fees of the students and the allotted share of the Treasury grant to university colleges are not sufficient without considerable additional support to carry on the higher education supplied by the college, the cost of which is constantly increasing. To make the work of the college fully effective, it is therefore desirable to obtain further endowment to the extent of 100,000l., or the equivalent income. The Senate of the University of London has shown approbation of the scheme for re