Committee's slow-action thermometers, in holes bored in the sides of the shafts.

The Committee desire to express their regret at the loss of their valuable member, Mr. Pengelly.

The Uniformity of Size of Pages of Scientific Societies' Publications.Report of the Committee, consisting of Professor S. P. THOMPSON (Chairman), Mr. G. H. BRYAN, Dr. C. V. BURTON, Mr. R. T. GLAZEBROOK, Dr. G. JOHNSTONE STONEY, and Mr. J. SWINBURNE (Secretary).

[PLATE I.]

THE importance of adopting one or two uniform standard sizes for the pages of scientific publications will be evident to all specialists who have collected reprints of papers on any branch of science, and who have endeavoured to have them bound into volumes. Such collections are of more than passing interest, and they might, with advantage in many cases, be handed down to posterity as records of work in any particular subject, and it is, therefore, of importance that they should not be spoiled by the omission of one or two papers whose size precludes them from being bound up with the rest.

The Committee have thought it advisable in their first year to confine their attention chiefly to reporting on the size of the pages of existing mathematical and physical publications, and to deciding on what sizes to recommend as standards. In the latter matter they have been largely guided by the consideration that uniformity has been already to some extent attained, and this report will show that the desired results can be accomplished without making any radical changes in the sizes of the principal journals, and, indeed, without altering most of them at all.

In deciding whether two papers can or cannot be bound together, the size of the margin is quite as important a factor as the size of the paper. Thus the Bulletin of the New York Mathematical Society' is more than a centimetre wider and two centimetres higher than the Report of the British Association,' and yet if it were cut down to the same size it would still have exactly the same margin at the sides, and 6 mm. more margin at the top and bottom of the pages. Again, the 'Proceedings of the Physical Society' are printed with the same type as the Philosophical Magazine,' although one is medium and the other demy octavo. Hence arises the necessity of taking an internal measurement of the space occupied by the letterpress, as well as an external measurement of the size of the paper page.

This may be estimated as in fig. 1, which represents the opened pages of a book. a, b denote the width and depth of a paper page, c is the distance from the outside edge of the letterpress to the back, and d is the distance from the top of the running headline or number of the page to the bottom of the last line of letterpress, exclusive of the 'signature.' Hence a-e is the margin at the side of the page, and b-d is the sum of the margin at the top and bottom, so that if these are equal each is equal to (b-d).

By adopting a minimum limit for the size of the pages (measured by a, b), and a somewhat smaller maximum limit for the internal measure

ments c, d, we shall secure that all papers can be bound together without cutting the margin down below a certain limit. In fixing the limits we

must allow for a little of the margin being cut away in binding.

OCTAVO PUBLICATIONS.-In the octavo sizes, the diagram, Plate I., shows an overwhelming preponderance of medium and demy octavo, the demy size being not only in the majority in point of number, but also including many of the most important publications. The royal octavo size recently proposed by the Royal Society is only repre

sented by about two journals, of which the 'Proceedings of the Royal Artillery Institution' is one. On the other hand, the space occupied by the letterpress, as shown by the measurements c, d, is no greater in several of the medium size than in the majority of demy octavo, and there would, therefore, be no difficulty in cutting these down in binding. The Committee, therefore, recommend the following sizes

Standard Octavo Size.--Paper demy, the pages measuring 14 cm. x 22 cm., or, when uncut, 53 in. x 83 in. The width c, measured from the stitching to the edge of the printed matter to be 12 cm., or 43 in., and the height, d, of the printed portion including the running headline, to be 18 cm.,or 7 in.

Limit of Octavo Size.-The paper page not to be less than 14 cm. × 21·5 cm., or 5 in. x 8 in., and the letterpress not to exceed the measurements c=12.5 cm., or 4 in., d=185 cm., or 74 in. Reprints and unbound numbers of journals to be issued with their edges uncut, or cut not more than 0.25 cm., orin., all round.

The use of standard as well as limiting sizes will easily be understood. Where publications fall within the limiting size there is little or no need for the size of the pages to be altered at present; but when any alteration is made in the size, or in the case of new journals or papers printed by their authors for private circulation, it would be desirable to conform exactly to the standard size, which would ultimately become general.

Taking, first, the limiting sizes, and allowing for 0-25 cm., orin., being cut off the margins in binding into volumes, the pages of these would measure 13.5 cm. × 21 cm., or 53 in. x 8 in., and this would allow of a margin of not less than 1 cm., orin., all round, which is quite enough.

If the standard sizes should become generally adopted the bound and cut volumes would measure the same, and the margin would be 2 cm., or in., at the sides, and 16 cm., or § in., at the top and bottom.

In the diagram it will be seen that there are journals which do not fall within the limiting dimensions. Where a and b fall short of the limits this could be remedied in some cases by leaving the edges uncut. Where c is too large, the letterpress could be brought a little nearer the stitching in imposing for press; where d is too great, the pages could be shortened by a line or two.

QUARTO PUBLICATIONS.-The corresponding dimensions for the prinquarto publications are given in the same diagram. Here, again, we edium quarto (24 cm. x 30 cm., or 9 in. x 12 in.) and demy quarto

(22 cm. x 28.5 cm., or 82 in. x 11 in.) almost universal, but the preponderating sizes approach most nearly to demy, and on account of the large margins of many journals this is the most convenient size of the two, besides making the volumes less unwieldy. The Committee, therefore, recommend, in cases where it is desired to retain the quarto size, the following measurements :

Standard Quarto Size.-Paper demy, the pages measuring, when uncut, 92 cm. x 28.5 cm., or 83 in. wide x 11 in. high. Reprints and unbound numbers of this size to be uncut, or cut 0.25 cm., orin. Measurements

of letterpress to be c=18.5 cm., or 74 in., d=21.5 cm., or 81⁄2 in.

Limits of Quarto Size.-Paper pages not to measure less than 21-5 cm., or 8 in., wide x 28 cm., or 11 in., high. Letterpress not to exceed the measurements c= =19 cm., or 7 in., d=23 cm., or 9 in.

The same remarks as to the advantage of standard and limiting sizes apply as in the case of the octavo. Allowing for 0.25 cm., orin., being cut off in binding, the limiting sizes will allow of a margin of not less than 2 cm., orin., all round, while the standard size will give a margin of 3.25 em., or 14 in., at the sides, and 2·5 cm., or 1 in., at the top and bottom.

PLATES often get sadly mutilated when different papers are bound together, and sometimes this even happens when a volume of any periodical is bound up. Where they are folded over they not infrequently get cut in two by the guillotine. To avoid this the Committee recommend that the dimensions of the illustrations should never exceed 13 cm. x 20 cm., or 54 in. x 7 in., for octavo plates, and 21 cm. x 25 cm., or 84 in. × 10 in., for quarto, the width being measured from the back of the book. Where plates have to be folded, the fold should be 12.5 cm., or 5 in., from the stitching in octavo, and 20.5 cm., or 8 in., in quarto papers. Any folding plate should, when referred to elsewhere than in the opposite page of letterpress, have a blank space equal to the breadth of the paper page at the left hand, so that when open it can be referred to without closing the portion of the book being read that refers to it. This should be carried out even when the diagram or plate would not otherwise have to be folded, in order to reduce the trouble of reference.

Each article should begin a page. If possible it should begin a righthand page. It is then possible to bind up any article with others on the same subject without having also to bind up the last half page of another paper. This difficulty can be overcome to some extent by splitting the paper. The pages of some of the journals abstracted in the 'Proceedings of the Physical Society' are split, one side being sent to each abstractor.

Comparison of Magnetic Standards.-Interim Report of the Committee, consisting of Professor A. W. RÜCKER (Chairman), Mr. W. WATSON (Secretary), Professor A. SCHUSTER, and Professor H. H. TURNER, appointed to confer with the Astronomer Royal and the Superintendents of other Observatories with reference to the Comparison of Magnetic Standards with a view of carrying out such Comparison.

PROFESSOR RÜCKER and Mr. Watson have carefully compared three Kewpattern magnetometers in order to investigate the causes of the discrepancies between the measurements of declination made with them. They

find that if the greatest care be taken in the manufacture of the wooden box and the metallic adjuncts which are close to the magnet the discrepancies disappear.

In other words, the cause of the difficulty, in these three instruments at all events, is, not the metal base, but the much smaller masses of metal which are nearer to the magnet.

The three magnetometers are now in good accord.

A week has been spent at each of four observatories for the purpose of comparing one of these magnetometers and a dip-circle with the observatory instruments. Professor Rücker made the observations at Kew and Falmouth; Mr. Watson, those at Stonyhurst and Valentia.

The greater part of the work which the Committee undertook has thus been accomplished.

It is still necessary to compare the instruments again with the instruments at Kew to ascertain that they are unaltered by transfer from one place to another; and as a new magnet-house is about to be built at Greenwich, it has been thought better to postpone the comparisons at that observatory until the house is ready for use.

The reductions of the observations which have been made are not yet finished. A full report will be made when the work is completed.

The Committee therefore ask to be reappointed, but no further grant is required.

The Application of Photography to the Elucidation of Meteorological Phenomena. Fifth Report of the Committee, consisting of Mr. G. J. SYMONS (Chairman), Professor R. MELDOLA, Mr. J. HOPKINSON, and Mr. A. W. CLAYDEN (Secretary). (Drawn up by the Secretary.)

IN the report which the Committee presented last year, it was proposed that an agreement should be entered into with the London and SouthWestern Railway Company for the use of a site on their land, in order to carry out some measurements of cloud altitudes by means of photography.

This has been done. The cameras have been placed in position, and almost the whole time at the disposal of the secretary for such purposes has been spent in perfecting the electrical connection for releasing the two shutters simultaneously. Considerable trouble has been experienced in doing this. The apparatus, which worked admirably over a short distance, proved unreliable over the greater distance (200 yards) at present adopted. The agreement with the railway company provides that the connecting wire shall be removed when not in actual use, thereby necessitating as light a wire as can be made to suffice, which of course implies a considerable resistance. The result is that a more sensitive electric detent is required for the shutters, especially as it seems not unlikely that the distance may have to be increased by another 100 yards when the measurement of the highest varieties of cloud is attempted. This is still engaging the attention of the secretary to the Committee.

Some observations have been made, but although they confirm the belief that the method will prove valuable they have not yet been reduced to actual measurements. It should be remembered that the method is only applicable to those varieties of cloud which are visible at the same time as the sun, and that the opportunities of making observations cannot