work. If the above names be adopted, they give us at once the foot-kinit as the unit of work based on the pound, foot, and second, the foot-pound (which varies with the value of g) being equal to g foot-kinits.

In like manner we have, for the metrical system, the metre-dyne and its derivatives.

But it would, I think, be advantageous to have short and independent names for these units. For, in the first place, we are thus saved from such cumbrous names as metre-kilodyne and metre-megadyne, which would be necessary in expressing large quantities of work; in the second place, energy of motion depends directly upon mass and velocity, and is only indirectly connected with the unit of force; and, in the third place, the characteristics of energy are such as specially entitle it to names suggestive of simplicity rather than of compositeness.

I propose, therefore, to call the foot-kinit, whether of work or energy, the erg. A thousand ergs to make one kilerg, which will be about 31 terrestrial foot-pounds, and a million ergs to make one poller, which is a little less than the work done by one horse-power in a minute.

The kinitic energy of m pounds, moving with a velocity of v feet per second, is m2 when expressed in ergs.

The energy value of a Fahrenheit unit of heat is 772 × 32·194 = 24,854 ergs.

=

In the metrical system, let the metre-dyne of work or energy be called the pone (from Tóvos). A thousand pones to make one kilopone, which is the work done by a kilodyne working through a metre, or by a dyne working through a kilometre, and is about I of the variable unit of work in common use among French 9.81 engineers, called the kilogrammetre. A million pones to make one megapone, which is about 723 terrestrial foot-pounds.

In employing the prefix mega to denote a million, I have followed the excellent example set by the B. A. Committee on Electrical Standards. As megerg would be intolerable, and megalerg sounds like a confusion of genders, I have substituted pollerg

In constructing a new nomenclature, the metrical system is entitled to the best names which can be found, but the pound and foot cannot be ignored. J. D. EVERETT

Rushmere, Malone Road, Belfast

Derivation of the Word "Britannia "

IF Mr. Edmonds considers himself right in his derivation of "Britannia" and "tin," he will have to explain on the same basis the conformable names, and this he will find difficult to do. The name B-ritannia corresponds with S-ardinia, D-ardania, and possibly with Mauritania, and these again with a number of river names of the root RDN (=RND, BRN, &c.), such as Rotanus, Rhodanus, Drinus, Eridanus, Artanus, Triton, Orethus, &c. B-radanus, P-rytanis, P-arthenias, V-artanus, are examples of B. Then there are K-artenus, I-ordanes, I-ardanes, I-ardenus. examples of Aternus, &c., Tanarus, &c., Mæander, &c., Orontes, &c. These must all be explained on one principle.

In the same way as Britannia is allied to river names, so are many of the ancient (classic) names of countries (except such as are volcanic) allied to river names of various roots, as RBD, &c., KKN, &C., SBN, &c.

These names are not explainable in Phœnician, because they were given long before the Phoenicians entered on the stage of history. They are Paleogeorgian, in a language to which Georgian, Lesghian, and other Caucasian languages are allied. These names were given by the Caucaso-Tibetans.

This is explained in my paper lately read before the Anthropological Institute and recorded in NATURE, and the name of Britannia is illustrated in papers sent in to the Society of Antiquaries and the Royal Irish Academy. HYDE CLARke

32, St. George's Square

Aurora by Daylight

THAT the Aurora Borealis has been seen by daylight has never been doubted by me, although till now I have not been able to collect sufficient evidence to induce others to believe in the possibility of it. Your correspondent Mr. John Langton, in your last issue, gives two instances of the aurora having been seen during day time, which, I think, ought to dispel all further doubt. However, to satisfy the most sceptical of your readers, the following few cases have occurred to me :

66

A. D. II22.

This same year died Ralph, Archbishop of Canterbury; that was on the 13th of the kalends of November (October 20). After this were many shipmen at sea and on the water, and they said that they saw on the northeast along the earth a great and broad fire, and it increased speedily upwards in extent towards the sky, and the sky opened itself in four parts and fought there against it as if it would extinguish it; but nevertheless the fire extended up to heaven. They saw that fire in the dawn of the day, and it continued until it was quite light. This was on the 7th of the ides of December (December 7).' Anglo-Saxon Chronicle.

It may seem bold to advance this as the record of an auroral appearance, but not to those who have studied this and other chronicles with their wearying vaguenesses. This passage gains clearness by the following lines from the "Prose Edda," concerning "The Twilight of the Gods and the Conflagration of the Universe," which I have elsewhere* supposed to be a description of the aurora borealis :

"The fire-reek rageth Around Time's nurse, And flickering flames

With heaven itself playeth."

In the "Second Continuation of the History of Croyland," there is the following curious passage, under A.D. 1467"For one day horsemen and men

in armour were seen rushing through the air; so much so, that St. George himself, conspicuous with the red cross, his usual ensign, and attended by a vast body of armed men, appeared visibly in great numbers. To show that we ought not to refuse our belief to what has been just mentioned, those persons to whom revelations of this nature were made were subjected to the most strict examination before the venerable Father Thomas, the Lord Archbishop of Canterbury."

I understand this occurrence to have taken place in the day between the rising and setting of the sun, because this passage is only part of a longer account of remarkable events which were said to have been observed in "one day.' I do not put this instance forward as one of very great value,† as the Chronicle of Ingulf is undoubtedly spurious, as shown by Dr. Hickes and Sir Francis Palgrave, but the continuation, I think, can be safely said to date about the end of the 15th or the beginning of the 16th century, which, if correct, will place the phenomenon above referred to amongst the earliest notices of daylight Auroras in English History, and will come next to that mentioned in the Anglo-Saxon Chronicle. Of course I only speak here of my own acquaintance with the Chronicles, there may be other records, but I have not had the opportunity of searching through every monastic production.

Leaving this field of speculation, I come next to a more reliable record. I give the whole of the passage, as it is not very long :"Aurora Borealis, seen in the Day-time at Canonmills." "The morning of Sunday, September 9, was rainy, with a light gale from the N.E. Before mid-day the wind began to veer to the west, and the clouds in the north-western horizon cleared away : the blue sky in that quarter assumed the form of a segment of a very large circle, with a well-defined line, the line above continuing dense, and covering the rest of the heavens. The centre

of the azure arch gradually inclined to the north, and reached an elevation of 20°. In a short time, very thin fleecy clouds began to rise from the horizon within the blue arch; and through these very faint perpendicular streaks of a sort of milky light could be perceived shooting; the eye being thus guided, could likewise detect the same pale streaks passing over the intense azure arch, but they were extremely slight and evanescent. Between nire and ten in the evening of the same day, the aurora borealis was very brilliant, so that there is no reason to doubt that the azure

and the fact that the contents of each pollen grain have to pass through the coats, both of the pollen tube and of the embryonic sack.' (I extract these latter addenda from Mr. Darwin's letter.) I do not much complain of having been sent on a false quest by ambiguous language, for I know how conscientious Mr. Darwin is in all he writes, how difficult it is to put thoughts into accurate speech, and, again, how words have conveyed false impressions on the simplest matters from the earliest times. Nay, even in that idyllic scene which Mr. Darwin has sketched of the first invention of language, awkward blunders must of necessity have often occurred. I refer to the passage in which he supposes some unusually wise, ape-like animal to have first thought of imitating the growl of a beast of prey so as to indicate to his fellow monkeys the nature of expected danger. For my part, I feel as if I had just been assisting at such a scene. As if, having heard my trusted leader utter a cry, not particularly well articulated, but to my ears more like that of a hyena than any other animal, and seeing none of my companions stir a step, I had, like a loyal member of the flock, dashed down a path of which I had happily caught sight, into the plain below, followed by the approving nods and kindly grunts of my wise and most-respected chief. And I now feel, after returning from my hard expedition, full of information that the suspected danger was a mistake, for there was no sign of a hyena anywhere in the neighbourhood. I am given to understand for the first time that my leader's cry had no reference to a hyena down in the plain, but to a leopard somewhere up in the trees; his throat had been a little out of order -that was all. Well, my labour has not been in vain; it is something to have established the fact that there are no hyenas in the plain, and I think I see my way to a good position for a look out for leopards among the branches of the trees. In the meantime, Vive Pangenesis. FRANCIS GALTON

The Hylobates Ape and Mankind

THE readers of Mr. Mivart's communication in NATURE for April 20, on the affinity of the Hylobates genus of ape to the human species, may be interested to learn that the fact was well known to the author of the Ramayana, the earliest Sanscrit epic, probably contemporaneous with the Iliad. In this poem the demigod Rama subdues the demon Ravana, and regains his ravished bride Sita by the assistance of a host of apes, which may be identified with Hylobates Hoolook. The human characteristics of these semi-apes, their gentleness, affection, good humour, sagacity, self-importance, impressionability, and proneness to melancholy, are portrayed with the most vivid strokes, and evidently from careful observation. See Miss Frederika Richardson's charming volume, "The Iliad of the East," a selection of legends drawn from the Ramayana. (Macmillan and Co., 1870.) April 27 R. G.

Tables of Prime Numbers

WHEN a number is given, and it is required, without the aid of tables, to find its factors, there is not, I believe, any other method known except the simple but laborious one of dividing it by every odd number until one is found that measures it, and if the number should be prime, this can only be proved by showing that it is not divisible by any odd number less than its square root. Thus to prove that 6966c07 is prime, it would be necessary to divide it by every odd number less than 2639, and even if a table of primes less than 2639 were at hand, about 380 divisions would be requisite.

On the other hand, there are few tables which are more easily constructed than tables of divisors, and it is the extreme facility of a systematic tabulation compared to the labour of isolated determinations, which has led to the construction of such elaborate tables on the subject as have been produced.

The principal tables are Chernac's, which give the factors of numbers from unity to a million; Burckhardt's, which extend as far as three millions, and Dase's, which form a continuation of Burckhardt's, and extend to ten millions.

The mode of formation of these tables was extremely simple. By successive additions, the multiples of 3, 5, 7, 11, 13, 17. were formed up to the limit to which the table was intended to extend; this gave all the numbers having these numbers for factors, and the primes were recognised from the fact of their not occurring as multiples of another prime less than themselves.

Practically the work was rendered even simpler by mechanical means; thus, forms were printed containing, say, a thousand

squares, and in these were written consecutive thousands of odd numbers in order; one number in each square, room being left for its divisors, if any, in the square. A pair of compasses was then taken and opened a distance corresponding to the prime whose multiples were to be obtained; for example, in marking the multiples of seven, the compasses were opened the width of seven squares, and then "stepped " along the lines starting from 7, thereby marking the numbers 7, 21, 35 and the number

7 was written in each of the squares in which a leg of the com. passes fell. When the factor was large it was more convenient to form a separate table of its multiples, and enter it in the square corresponding to the latter. Many simplifications were introduced in the details of the construction; for instance, Burckhardt had a copper plate engraved with 77 (=7×11) squares one way and So the other; by this arrangement the multiples 7 and 11, which were of the most frequent occurrence (for all multiples of 2, 3, and 5 were rejected from the tables), occupied the same place on each sheet, and he was thus enabled to engrave the numbers 7 and 11 on the plate, so that these numbers were printed in all the squares containing the numbers they measured.

Dase, who originally applied himself to the construction of the tables at the suggestion of Gauss, left behind him in manuscript at the time of his death, in 1862, the seventh and part of the eighth million complete, besides a considerable portion of the ninth and tenth millions. The seventh, eighth, and ninth millions were completed by Dr. Rosenberg, and published by a committee at Hamburgh. In the preface to the ninth million (1865), which is the last I have seen, it is stated that the tenth million, which was nearly ready, was the last the committee intended to publish.

My object in writing this letter is not only to call attention to a most valuable series of tables, which seem to have scarcely excited so much interest as they deserve, but also to ask if any of your readers can inform me if the work is being continued, or if there is any chance of its continuation. It is not often that tables are so indispensable as in the present case, or that a want so pressing can be supplied with such comparative ease; and the cessation of the tables would be a real calamity. The tenth million has, I presume, been published.

At the British Association Meeting at Dundee in 1867, a list of 5.500 large prime numbers was communicated to Section A by Mr. Barrett Davis. A short discussion took place on the "reading" of the paper, in the course of which it was stated that Mr. Davis's table was unaccompanied by any explanation of how the numbers had been obtained, or on what grounds they were asserted to be prime; it was also asserted that Mr. Davis wished to keep his method secret.

Perhaps some reader of NATURE can say whether Mr. Davis's numbers have been printed. If they exceed Dase's limit, their publication (if they have not yet been published) is very desirable; and even supposing they are given in Dase's tables, it would be valuable to know how far the latter have been verified by them. The statement about Mr. Davis's method being secret was probably founded on some mistake, and no doubt Mr. Davis would not object to explain it. J. W. L. GLAISHER Trinity College, Cambridge, April 29

Units of Force and Energy

There

THE best root for the name of a unit of force is dúvaus. is, therefore, no ground for Mr. Muir's complaint (NATURE, vol. iii. p. 426), and I now venture to propose that the name dyne be given to that force which, acting on a gramme for a second, generates a velocity of a metre per second. A thousand dynes to make one kilodyne, and a million dynes one megadyne.

Borrowing a hint from Mr. Muir, I would point out that the kilodyne may also be defined as the force which, acting on a kilogramme for a second, generates the velocity of a metre per second, or, as the force which, acting on a gramme for a second, generates a velocity of a kulomere per second.

The kinit, or pound-foot-second unit of force, is about 1381 dynes. Very roughly expressed in terrestrial gravitation measure, the kinit is the gravitating force of half an ounce, the dyne of about 14 grains, the kilodyne of about of a pound, and the megadyne of 2 cwt., the approximation being much closer in this last case than in the others, so that within one part in 400 we have 10 megadynes the force of terrestrial gravity on a ton. I have often felt the want of a name for an absolute unit of energy, or, what amounts to the same thing, an absolute unit of

work. If the above names be adopted, they give us at once the foot-kinit as the unit of work based on the pound, foot, and second, the foot-pound (which varies with the value of g) being equal to g foot-kinits.

In like manner we have, for the metrical system, the metre-dyne and its derivatives.

But it would, I think, be advantageous to have short and independent names for these units. For, in the first place, we are thus saved from such cumbrous names as metre-kilodyne and metre-megadyne, which would be necessary in expressing large quantities of work; in the second place, energy of motion depends directly upon mass and velocity, and is only indirectly connected with the unit of force; and, in the third place, the characteristics of energy are such as specially entitle it to names suggestive of simplicity rather than of compositeness.

I propose, therefore, to call the foot-kinit, whether of work or energy, the erg. A thousand ergs to make one kilerg, which will be about 31 terrestrial foot-pounds, and a million ergs to make one pollerg, which is a little less than the work done by one horse-power in a minute.

The kinitic energy of m pounds, moving with a velocity of v feet per second, ismo2 when expressed in ergs.

=

The energy value of a Fahrenheit unit of heat is 772 × 32 194 24,854 ergs.

In the metrical system, let the metre-dyne of work or energy be called the pone (from óvos). A thousand pones to make one kilopone, which is the work done by a kilodyne working through a metre, or by a dyne working through a kilometre, and is about I of the variable unit of work in common use among French engineers, called the kilogrammetre. A million pones to make one megapone, which is about 723 terrestrial foot-pounds.

9.81

In employing the prefix mega to denote a million, I have followed the excellent example set by the B. A. Committee on Electrical Standards. As megerg would be intolerable, and megalerg sounds like a confusion of genders, I have substituted pollerg.

In constructing a new nomenclature, the metrical system is entitled to the best names which can be found, but the pound and foot cannot be ignored. J. D. EVERETT

Rushmere, Malone Road, Belfast

Derivation of the Word "Britannia "

IF Mr. Edmonds considers himself right in his derivation of "Britannia" and "tin," he will have to explain on the same basis the conformable names, and this he will find difficult to do. The name B-ritannia corresponds with S-ardinia, D-ardania, and possibly with Mauritania, and these again with a number of river names of the root RDN (=RND, BRN, &c.), such as Rotanus, Rhodanus, Drinus, Eridanus, Artanus, Triton, Orethus, &c. B-radanus, P-rytanis, P-arthenias, V-artanus, are examples of B. K-artenus, I-ordanes, I-ardanes, I-ardenus. Then there are examples of Aternus, &c., Tanarus, &c., Mæander, &c., Orontes, &c. These must all be explained on one principle.

In the same way as Britannia is allied to river names, so are many of the ancient (classic) names of countries (except such as are volcanic) allied to river names of various roots, as RBD, &c., KKN, &C., SBN, &c.

These names are not explainable in Phoenician, because they were given long before the Phoenicians entered on the stage of history. They are Paleogeorgian, in a language to which Georgian, Lesghian, and other Caucasian languages are allied. These names were given by the Caucaso-Tibetans.

This is explained in my paper lately read before the Anthropological Institute and recorded in NATURE, and the name of Britannia is illustrated in papers sent in to the Society of Antiquaries and the Royal Irish Academy. HYDE CLARKE

32, St. George's Square

Aurora by Daylight

THAT the Aurora Borealis has been seen by daylight has never been doubted by me, although till now I have not been able to collect sufficient evidence to induce others to believe in the possibility of it. Your correspondent Mr. John Langton, in your last issue, gives two instances of the aurora having been seen during day time, which, I think, ought to dispel all further doubt. However, to satisfy the most sceptical of your readers, the following few cases have occurred to me :-— "A. D. 1122. This same year died Ralph, Archbishop of Canterbury; that was on the 13th of the kalends of November (October 20). After this were many shipmen at sea and on the water, and they said that they saw on the northeast along the earth a great and broad fire, and it increased speedily upwards in extent towards the sky, and the sky opened itself in four parts and fought there against it as if it would extinguish it but nevertheless the fire extended up to heaven. They saw that fire in the dawn of the day, and it continued until it was quite light. This was on the 7th of the ides of December (December 7).' Anglo-Saxon Chronicle.

It may seem bold to advance this as the record of an auroral appearance, but not to those who have studied this and other chronicles with their wearying vaguenesses. This passage gains clearness by the following lines from the "Prose Edda," concerning "The Twilight of the Gods and the Conflagration of the Universe," which I have elsewhere* supposed to be a description of the aurora borealis :

I understand this occurrence to have taken place in the day between the rising and setting of the sun, because this passage is only part of a longer account of remarkable events which were said to have been observed in "one day." I do not put this instance forward as one of very great value,† as the Chronicle of Ingulf is undoubtedly spurious, as shown by Dr. Hickes and Sir Francis Palgrave, but the continuation, I think, can be safely said to date about the end of the 15th or the beginning of the 16th century, which, if correct, will place the phenomenon above referred to amongst the earliest notices of daylight Auroras in English History, and will come next to that mentioned in the Anglo-Saxon Chronicle. Of course I only speak here of my own acquaintance with the Chronicles, there may be other records, but I have not had the opportunity of searching through every monastic production.

"The

Leaving this field of speculation, I come next to a more reliable record. I give the whole of the passage, as it is not very long "Aurora Borealis, seen in the Day-time at Canonmills." morning of Sunday, September 9, was rainy, with a light gale from the N. E. Before mid-day the wind began to veer to the west, and the clouds in the north-western horizon cleared away: the blue sky in that quarter assumed the form of a segment of a very large circle, with a well-defined line, the line above continuing dense, and covering the rest of the heavens. The centre of the azure arch gradually inclined to the north, and reached an elevation of 20°. In a short time, very thin fleecy clouds began to rise from the horizon within the blue arch; and through these very faint perpendicular streaks of a sort of milky light could be perceived shooting; the eye being thus guided, could likewise detect the same pale streaks passing over the intense azure arch, but they were extremely slight and evanescent. Between nire and ten in the evening of the same day, the aurora borealis was very brilliant, so that there is no reason to doubt that the azure * Vide NATURE, vol. iii. p. 175.

For a similar case to this see note to my letter on the aurora borealis in NATURE, vol. iii., p. 487.

I have just been informed by a friend whose veracity I would be the last to question, that he saw a very faint arch in the castern sky on the afternoon of the 10th inst. (about 4.30 P.M.). There were no clouds near it, while the background was a beautiful azure. The colour of the arch was of a much fainter blue, or, as he calls it, "a whitish blue," and was almost a perfect semicircle. I have not the least doubt that it was a daylight' aurora; it must be remembered that on the previous night there was a most magnificent aurora borealis.

66

In conclusion, after carefully examining the facts contained in the various communications to your journal, as well as those which I have collected, I cannot see any reason for doubting the possibility of the aurora borealis being seen by daylight. It will be interesting to know what those daylight phenomena are, if not JOHN JEREMIAH

auroras.

Red Lion Street

The Irish Fern in Cornwall YOUR correspondent having, much to my regret, so exactly informed the ruthless collectors" where they are to look for this fern, I fear that after the ensuing autumnal ravages not a single frond will be left to speak for itself. Permit me, therefore, to state that the fern unquestionably grows, or did grow, at the place indicated, and was, I believe, first recognised in 1866 by Mr. Robert Were Fox, F.R.S., who has a plant he thus obtained still growing in his fernery at Penjerrick near this

The Prevalence of West Winds

IN a letter with this heading in NATURE for February 16th, Mr. Murphy has very roundly objected to certain views which I have put forward regarding the predominance of westerly winds. In the paper read before the British Association, to the abstract of which he refers, and which was itself little more than a résumé of the propositions maintained at greater length in my "Physical Geography," reviewed by "A. B." in NATURE for March 16th, my object was not so much to show that westerly winds predominated in volume over easterly winds, as to show that all prevailing winds, not westerly, may be properly considered as deflected or secondary currents of air, and that more especially the trade winds may be so considered. I have supported this view by a detailed examination of the geographical circumstances, habitudes, and characteristics of the principal winds; but to have included every local exception-as "A. B." seems to consider I ought to have done-would have required more time than even the most industrious can spare, an amount of special topographical knowledge which is practically unattainable, and would have had no important bearing on the main question. I may go even further. I may say that, from a general point of view, isolated local registers have no value at all, unless the method of observing and the position of the vane are distinctly made known. It would be perfectly easy to name a dozen localities in Wales, in the Lake District, or in Scotland, where a vane would show a prevailing wind widely different from the W.S. W., which, however, we have no difficulty in accepting as the prevailing wind of the country; even at Liverpool the prevailing wind has been observed to be W. N. W., and at Valentia there is a marked difference between the wind in the northern and southern entrance. In Mr. Buchan's paper in the Transactions of the Royal Society of Edinburgh, December, 1869, I find that at Irkutsk the wind is almost always due north, or due south, would "A. B." imply that the Irkutsk observations afford any information as to the prevailing wind of Siberia?

In another paragraph, "A. B." considers that the preponderance of westerly winds cannot be very great. So far as the area over which westerly winds blow is concerned, I would partly agree with him; taking into account the constant interruptions to the west winds in the temperate zones, and on the other hand their frequent intrusion into latitudes considerably below 30°, more especially in the Pacific, and their prevalence during several months of the year over a large portion of the Indian Ocean, I am inclined to reckon the ratio of the area of westerly winds to the area of easterly winds as approximately 13: 10. But such an estimate

Jameson's Journal, quoted in the "Arcana of Science and Art" for 1828.

does not in any way include the velocity of the wind; and since the velocity of the west winds of temperate latitudes is, in the mean, about double that of the easterly winds of tropical, it would follow that the respective volumes of the winds bear to each other a much larger ratio, which, allowing freely for every reasonable reduction, cannot be less than 2: I. And this estimate still relates only to the lower strata of the atmosphere, through a height probably not exceeding 12,000 feet. Our knowledge of the winds above that height is very limited; but since, wherever observation extends, it points out to us a strong, frequently even a violent west wind, it seems to me that we have a fairly presumptive proof that the prevailing direction of the upper current is from the west. I base this belief entirely on the evidence which we have, defective as it is and as it almost necessarily must be; to explain the fact by a reference to a difference of barometric pressures, concerning which we have positively no evidence at all, is a task which I most willingly leave to my reviewer. But if, as I have maintained, we may fairly assume that the upper current has an almost invariable direction from the west, and that too with a comparatively high velocity, the ratio of the volumes of westerly and easterly winds is enormously increased, and if the upper part of the air, being quite half of the whole, is moving from the west with a mean velocity of 40 miles an hour, then, as we have already taken 20 miles, or the velocity of the trade winds, as the standard or unit of reference, we have the ratio of westerly to easterly winds as about 6: 1.

The question which Mr. Murphy has suggested no doubt here arises: Must not this preponderance of westerly winds affect the rotation of the earth? I have throughout maintained the existence of this preponderance solely by geographical proof, and conceiving that the evidence is conclusive, whilst no meteorological theory points to any explanation of it, I am compelled to attribute it to the action of some force external to the earth; possibly, as I have endeavoured to show, to the attraction of the sun, moon, and other heavenly bodies; possibly also to some other force, magnetic or meteoric, of whose action we have as yet no knowledge or understanding: but supposing, as I do, that the force which produces this motion is external to the earth, it is impossible to avoid the conclusion that it does tend to increase the earth's velocity of rotation. On the other hand, there are forces, admitted by all naturalists, in constant action, which tend to decrease the velocity of rotation; and a certain amount of wonder that the decrease so caused is so small as observation proves it to be is implied, rather than expressed, in our most valuable works on Natural Philosophy. If it is impossible in the present state of our knowledge to show exactly what such decrease is and ought to be, it is certainly impossible to say that it is not to some extent counterbalanced by a contrary tendency towards an increase, such as I have shown probably exists. At any rate, I know of nothing connected with the rotation of the earth which in any way controverts or affirms the proposition which I have put forward, based on geographical evidence only.

I had written this before seeing Mr. Murphy's second letter on the subject in NATURE for March 30, but as he has in it merely repeated his former arguments, it is unnecessary to notice it more particularly. J. K. LAUGHTON

Royal Naval College, Portsmouth

SUBMARINE TELEGRAPHS

IT may possibly be within the memory of some persons that, about the year 1840, Sir C. Wheatstone first conceived the idea of transmitting messages under the sea, and practically carried out at that time the first submarine telegraph cable. Selecting Swansea Bay, South Wales, as the chosen spot for his experiment, the great inventor sat in an open boat, about three miles from the Mumbles Lighthouse, with the lighthouse keeper as his assistant. A conducting wire, insulated with hemp and a resinous compound, served as the electric communication between his open boat and the shore. It is from the successful results of this first crude experiment, and Wheatstone's investigations into the laws that regulate the transmission of electric currents through metallic conductors, published shortly afterwards in the Philosophical Transactions of the Royal Society of London, that our present system of the testing of submarine cables is based,