Page images

A short history is given in chapter viii. of the general development of triangulation from the time of Snellins to the present day, but it consists mainly in mentioning some of the more striking incidents connected with the subject, such, for instance, as the use of electric light in the work connecting Spain and Algiers.

The question of lateral refraction is gone very fully into, and a table is given showing the mean triangular errors proportional to triangles of different sizes. The deduction is then made that the effect of lateral refraction increases with the mean length of a side of a triangle up to about 90 kilometres, after which it begins to decrease again. This is a particularly interesting problem, and the table, which is taken from a recent triangulation in Germany, certainly confirms the deduction. It is of course assumed that the closing errors of triangles are due in most part to lateral refraction. The usual methods of calculating and computing triangulation are very thoroughly dealt with.


Precise levelling forms the greater part chapter ix., but an exceedingly full and clear demonstration is given at the beginning of the various formulæ relative to atmospheric refraction and to trigonometrical differences in heights of stations. This is treated in a really very clear manner.

Three chapters are devoted to the theory of probability and its application to geodesy. The theoretical portion has been well demonstrated, and differs but little from the numerous text-books on this subject; but where the attempt is made to apply the method of least squares to a network of triangulation, the want of taking a practical example is at once felt. Clarke, in his "* Geodesy," gives numerous examples of how to apply theory to practice, but Signor Pizzetti leaves the student utterly in the dark on this important point.

Perhaps the two most interesting chapters are left to the end. They deal with the subject of projections, which is gone into with every care. There is scarcely any well known projection which is not very fully explained.

Altogether this book is a distinct addition to any geodetic library, W. J. JOHNSTON.


The Food Inspector's Handbook. By Francis Vacher. Fourth edition. Pp. xvi+231; illustrated. (London: The Sanitary Publishing Co., 1905.) Price 35. 6d. net.

THIS is a pocket volume intended for the use of sanitary and other officers concerned in the inspection of food. It describes, in simple, untechnical language, the naked-eye characters of the various foodstuffs met with in ordinary commerce, and points

out the physical signs by which unwholesome food may be detected.

The first sixty pages deal chiefly with the statutory powers by virtue of which the food-supply of the community is supervised. They include a summary, with explanatory comments, of the various enactments --Public Health Act, Sale of Food and Drugs Acts, and so on-bearing upon the control of food from the inspector's point of view. Next follow chapters treat

ing of meat, poultry, and fish. This is the most important part of the book, and the notes upon the ante- and post-mortem signs of those diseases which render flesh-food unfit for consumption, or which seriously depreciate its quality, will be especially valuable to sanitary officers who have had no veterinary experience. A subsequent chapter is allotted to fruit and vegetables, and one to milk; the rest of the book gives short descriptions of cereals, dairy products, tea, sugar, spices, and so forth. This section, though of interest to the food inspector, is of less concern to him than the foregoing, the quality of the articles mentioned being generally a matter for decision by analysis, not for condemnation at sight.

The author gives sensible advice, and his little volume should be found very useful to those for whom

it is written. The only inaccuracy we have noted is suggested in the statement that "Dutch cheese is below the standard per cent. as regards fat "; this might imply that there is a legal standard, which is not the fact. C. S. Elettricista. By Attilio (Milan Ulrico Hoepli,

Manuale dell'Ingegnere Marro. Pp. xv+689. 1905.) Price 7.50 lire.

THIS book forms one of the useful series of "Manuali Hoepli," which already comprises over 800 distinct treatises. Its aim is to give to engineers and electrical constructors most of the information and data that they are likely to require in practice. On this account it is not so much a text-book as a classified collection of rules and data; but on account of its containing a large amount of explanatory matter it lies intermediate between a treatise on electrical

engineering and a pocket book of electrical rules and tables. The type being small but clear, a very large amount of useful information is collected in a small compass. The numerical data have been obtained principally from recent papers published in the journals of electrical engineering, and are collected in 115 tables. The work is illustrated with 192 cuts and is furnished with a good index. Its size well adapts it to be a handy pocket book of reference, and it is likely to prove of considerable use.

Poisonous Plants of all Countries. By A. B. Smith. Pp. xvi+88. (Bristol: J. Wright and Co., 1905.) Price 2s. 6d. net.

THE author has collected a fairly representative list of poisonous plants, which he has arranged according to the action produced and the organs affected, but there is no mention made of the part or parts of the plant which furnish the poison. The descriptions, which form the main part of the text, are sufficient where reference is made to the whole plant, but the majority are too meagre to be diagnostic. The string of vernacular names which is quoted in several cases does not serve any useful purpose, whereas beyond the mere name of the toxic principle information which is much required is not given.

[blocks in formation]

as great as that deduced from Planck's formula for this case. In connection with similar work of his own, Mr. Jeans (Phil. Mag., July) has just pointed out that I have introduced a redundant factor 8 by counting negative as well as positive values of my integers §, n, S. I hasten to admit the justice of this correction. But while the precise agreement of results in the case of very long waves is satisfactory so far as it goes, it does not satisfy the wish expressed in my former letter for a comparison of processes. In the application to waves that are not long, there must be some limitation on the principle of equi-partition. Is there any affinity in this respect between the ideas of Prof. Planck and those of Mr. Jeans? Terling Place, Witham, July 7. RAYLEIGH.

Proposed Observation of Mercury during the Solar Eclipse.

DURING the eclipse of the sun on August 30 next there will be an opportunity of making a very interesting observation on the planet Mercury, to which I ask permission to direct attention.

Mercury at the time of the eclipse will be very close to the line joining the earth and sun-about 2° 54′ south and 2° 54' preceding the sun-i.e. at an angular distance from the sun's centre of nearly 4° 6'. Accordingly, the illuminated part of Mercury will be an excessively thin crescent which, if Mercury have an atmosphere, will have its horns prolonged by atmospheric refraction.


If a sufficiently skilled observer is provided with telescope upon which he can use a power of 200 without loss of definition, and mounted-probably as an altazimuth-so that it can be set beforehand upon Mercury, the apparent size of Mercury will be that which would be presented by a circle one-tenth of an inch across, viewed with the naked eye from a distance of ten inches.

This ought to be sufficient magnification to see whether the horns of the crescent are prolonged, and, if so, it is perhaps not impossible that the light would be sufficient to enable a spectrum of the tips of the crescent to be


If the whole of this programme can be carried out, we should find out whether Mercury has an atmosphere, and possibly learn something as to the constituents of the atmosphere. G. JOHNSTONE STONEY.

30 Ledbury Road, W., July 10.

The Planet Uranus.

ASTRONOMICAL amateurs will have an excellent opportunity of identifying Uranus on about July 16 next, for the planet will be in conjunction with the star 1 Sagittarii (mag. 5.3) on the night following that date.

The position of Uranus at transit (10h. 30m.) will be:R.A. 18h. 5m. 58s., Dec. S. 23° 42' 21", while that of the star will be (1905-5):


R.A. 18h. 5m. 57s., Dec. S. 23° 43′ 16′′,

so that the planet will pass about 1 minute of arc north of the star. The latter may be easily picked up, as it is nearly 4° S. of the triple star μ or 13 Sagittarii (mag. 4.1).

Unfortunately, the objects will be low in altitude (15°), and the moon happens to be full on the date of conjunction. On June 24 and July 8 I found Uranus a little fainter than the star 1 Sagittarii. I have carefully observed the planet on several nights in a 12in. Calver reflector, powers 100 to 475. The disc appeared faint with a bluish tinge, and no belts or other markings could be detected, but the telescope is too small to deal effectively with an object of this description. Bristol, July 9.


The Exploration of the Atmosphere above the Atlantic.

steamer, is now being partially realised. Last summer Prof. Hergesell, on board the Prince of Monaco's steamyacht Princess Alice, executed sixteen kite-flights above that part of the Atlantic bounded by Spain, the Canaries and the Azores, but without finding the expected southwest anti-trade, although a height of nearly 15,000 feet was reached (NATURE, vol. lxxi. p. 407). The present expedition, which will repeat Prof. Hergesell's investigations and continue them further south, is made possible through the cooperation of our distinguished French colleague, M. Teisserenc de Bort, whose steam-yacht Otaria, of 350 tons, with a speed of 11 knots, and fully equipped for aerial exploration, has just sailed from Gibraltar, and, at the joint expense of her owner and the writer, will proceed towards the equator by way of Madeira, Canary and Cape Verde Islands, making frequent soundings with kites through the trade winds and equatorial calms. By means of the self-recording instruments lifted by the kites, it is expected that there will be ascertained the thermal and hygrometric conditions of the various strata traversed, and the depth and force of the trade wind in the different latitudes. If the kites do not reach the south-west return trade, which has been observed on the Peak of Teneriffe, the vertical range of observation may be increased by liberating hydrogen balloons from Madeira and noting

their drift.

Mr. Clayton, meteorologist of the Blue Hill Observatory, left Boston on June 3 to join the Otaria at Gibraltar. During his voyage to the Azores on the White Star liner Romanic he flew kites, with instruments attached, almost daily to the height of from five-eighths to three-quarters of a mile, thus securing the highest observations in midAtlantic, and it is interesting that this was done on June 7, the day appointed for the international observations in the upper air here at Blue Hill and in Europe. In general, the temperature was found to decrease with altitude at less than the adiabatic rate, and the relative humidity to decrease also, but in one ascent there was a rise of temperature with altitude, preceding a change of wind from west to south. A new form of folding kite was employed, and it is encouraging to learn that the heights attained were limited by the length and strength of the wire on the hand-reel, which did not permit more than one of these kites to be attached. On the two days when no flights were made, a following wind became too light on board the steamer to lift the kites, whereas, on the yacht, this condition would have been obviated by simply lying-to, or steaming against the wind.

While the cruise of the Otaria, which is to last only six weeks, can hardly do more than elucidate certain questions relating to the high atmosphere in the tropics, it will demonstrate the possibilities and difficulties attending the extensive survey that the writer desires to undertake, and which received the endorsement of the International Meteorological Committee at Southport in 1903. A. LAWRENCE ROTCH.



Blue Hill Meteorological Observatory, Hyde Park, Mass., U.S.A., June 26.

Ancient Antarctica.

PROF. H. F. OSBORN has said that the demonstration "the former existence of an Antarctic continent one of the greatest triumphs of modern science." But even if this be true, everyone must allow that it occurred a very long time ago. This is proved by the great differences that exist between the floras and faunas of the three great southern continents. These differences are much greater than those between the floras and faunas of North America and Eurasia, and consequently the land connections must have been broken up in the south long before they were in the north. We infer the former existence of an Antarctic continent from the existence of granite and foliated schists in South Victoria Land, and evidence that it was formerly connected with northern lands is found in the existence of flightless insects living there in the few patches of mosses and lichens which manage to struggle through the winter. These insects are not flightless through degener

A PLAN for systematic work of this kind, which was proposed by the writer in 1901 at the Glasgow meeting of the British Association (Report, p. 724) after he had obtained the first observations with kites flown from a Transatlantication, but belong to an order which never possessed wings.

It is very improbable that the ancestors of these minute insects were carried or blown to where they are now found; they must have travelled to their present positions by land. That is, the Antarctic continent south of New Zealand and Patagonia must, at some time or other, have joined on to northern lands.

In the islands of the Antarctic Ocean we have further evidence of a former land connection in the earthworms belonging to the family Acanthodrilidæ, which are characteristic of Antarctic regions. A spider also lives on Bounty Islands which is closely related to one from Cape Horn. But spiders seem to have special facilities for crossing barriers, and the insects found on Bounty Islands are all related to New Zealand forms. I do not include here the evidence of the plants of the Antarctic islands, for most plants do not require that the land should be actually continuous to enable them to spread.

But if the flightless insects and the earthworms imply a former connection with northern lands, that connection must have been a very long time ago, before the spread of insects and angiospermous plants over the world, that is, not later than the Jurassic period. If there had been any land connection in Tertiary times, there would have been a much greater mixing of the animals and plants.

It is evident that the flora, and perhaps the fauna, of Antarctica were formerly much richer than at present, as is proved by the fossil plants of South Victoria Land, and it is also probable that both fauna and flora were killed off by an increasingly rigorous climate. It is not necessary to assume a former Glacial epoch for this, for higher plants and animals could hardly resist the present climate, and there is no palæontological evidence of a period of greater cold than now having ever existed in the southern hemisphere. On the contrary, the biological as well as the paleontological evidence is against the idea. For the much modified plover, Chionis, and the insects of Kerguelen Land, as well as the remarkable flora of the Antarctic islands, show that the islands could not have been covered with ice for a very long time.

The relations between the avifaunas of Australia and South Africa are much closer than exist between those of Australia and South America, and this is just what we should expect if the ancestors of the present birds had spread down from the north under the present condition of land and sea, for the land connection between Australia and South Africa is far more intimate than that between the former place and South America. But the contrary is the case with the Mammalia, some of the tortoises, snakes, frogs, some of the fresh-water fishes, a large number of insects, and the family Cryptodrilidæ of earthworms. This implies that at some former time a closer connection existed between Australia and South America than between Australia and Africa. The question is, Was this connection by means of an Antarctic continent? Or was it by a Pacific continent?


The principal objection to the southern route is that the connection between Australia and South America is shown by a number of subtropical animals-such Osteoglossum and Ceratodus-none of which have left any trace of their passage through New Zealand. We cannot suppose that New Zealand was disconnected at the time from the Antarctic continent, for it, also, has distinct relations with South America, but for the most part by means of different animals from those which show the Australian connection. If the connection was in either the Cretaceous or the Eocene period, we might suppose that the climate was warm enough for the passage of the subtropical animals by the Antarctic route, but, if so, why are there no traces of marsupials and South American frogs in New Zealand? If, on the other hand, we suppose the ancestors of these animals to have crossed from Australia to South America by a South Pacific continent, we can understand how the subtropical forms would not have come so far south as New Zealand, while the New Zealand forms would have crossed at a higher latitude. In favour of this we have a member of the Iguanidæ in Fiji, as well as the evidence of the land shells of Polynesia, which are not a collection of waifs and strays, but form a distinct group of a very early type, which, however, has not yet been found in South America.

We still have to consider the floras and the marine faunas of the Antarctic islands. Here we see a number of birds such as cormorants and gulls-as well as fishes and plants, which could hardly spread round the world under the present conditions of land and water. That this spreading was a comparatively late one is proved by the near relations between the species. But if there had been continuous land at the time, land animals would have spread with the marine ones. It is therefore necessary to suppose that this last spreading of species in Antarctic latitudes was by means of a number of islands. Probably this was in Pliocene times, if we may judge by the amount of differentiation which has taken place since then.

I therefore conclude that the hypothesis which best explains the phenomena is the following:

(1) That in the Jurassic period an Antarctic continent existed which connected South America with New Zealand and South Africa.

(2) That this continent sank in the Cretaceous period, and that Antarctica has never since been connected with northern lands.

(3) That in the Cretaceous or early Eocene a Pacific continent connected New Guinea and New Zealand with Chili.

(4) That this land sank at the close of the Eocene. (5) That in the Pliocene a number of islands existed in the Antarctic Ocean, which have since then disappeared. F. W. HUTTON.

[ocr errors]

The British Slugs.

MR. J. W. TAYLOR has just published part ii. of his admirable Monograph of the Land and Freshwater Mollusca of the British Isles," containing a discussion of the slugs of the genus Arion. It is a matter of interest that, notwithstanding the great amount of information gathered in recent years, the beautiful bicoloured varieties of A. ater appear to hold their own as truly endemic inhabitants of Britain. These are three in number, though Mr. Taylor treats the third as merely a subvariety.

(1) Arion ater, var. albolateralis, Roebuck, 1883. Back black, sides white.

(2) Arion ater, var Roebucki (bicolour, Roebuck, in error). Back brown, sides yellow.

(3) Arion ater, var. Scharffi, Cockerell, 1893. Back black, sides yellow.

Mr. Taylor retains the name bicolour for the second variety, but it is not the bicolour of Moquin-Tandon, as was formerly supposed, and a new name is necessary. It is appropriate to name it after Mr. Roebuck, who first made it known. These magnificent slugs are of western distribution in the British Islands, and have quite a wide range. The only evidence of their occurrence on the Continent is Scharff's statement that Simroth found specimens similar" to var. Roebucki on the shores of the Baltic; and the possibility that the Norwegian var. medius, Jensen, may be similar to albolateralis, though it is very likely not even of the same species. A quite different variety of A. ater is the wholly black form aterrima, said to be especially northern and montane. According to Mr. Taylor, this is exclusively British, except that it appears to be represented in Spain and Portugal by a similar animal named hispanicus by Simroth. However, I had always regarded this aterrima variety as the one so described from France by Dumont and Mortillet (cf. Science Gossip, 1889, p. 212, "the pitchy black variety found in swamps "), and if it is not, the name aterrima, applied to it by Mr. Taylor, cannot stand. At the opposite pole of variation from aterrima is the brilliant red form A. ater, var. coccinea (Gistel), which is hardly ever found in England, but is abundant in the warmer and drier regions of Central Europe.

Incidentally, it may be remarked that the name Arion hortensis, var. subfusca, employed by Mr. Taylor, cannot be retained, as it is founded on Limax subfuscus, C. Pfr., a homonym of L. subfuscus, Draparnaud.

University of Colorado, June 26.


[ocr errors]



OME years ago I referred in NATURE to the numerous alignments of stones in Brittany, and I was allowed by Lieut. Devoir, of the French Navy, to give some of his theodolite observations of the directions along which the stones had been set up. The conclusion was that we were really dealing with monuments connected with the worship of the sun of the May year, a year which the recent evidence has shown to have been the first recognised after the length of the year had been determined; thus replacing the lunar unit of time which was in vogue previously, and the use of which is brought home to us by the reputed ages of Methuselah and other biblical personages, who knew no other

measurer of time than the moon.

There was also evidence to the effect that in later times solstitial alignments had been added, so that the idea that we were dealing with astronomically oriented rows of stones was greatly strengthened, not to say established.

So long as the Brittany alignments were things of mystery, their origin, as well as that of the more or less similar monuments in Britain, was variously explained; they were models in stone of armies in battle array, or they represented funeral processions, to mention only two suggestions. I should add that Mr. H. Worth, who has devoted much time to their study, considers that some sepulchral interest attaches to them, though he thinks it may be argued that that was secondary, even as are interments in cathedrals and churches. About burials associated with them, of course, there is no question, for the kistvaens and cairns are there; but my observations suggest that they were added long after the avenues were built, as some cairns block avenues. Perhaps a careful study of the mode of burial may throw light on this point.

The equivalents of the Brittany alignments are not common in Britain; they exist in the greatest number on moor, whither I went recently to study them. conditions on high Dartmoor are peculiar.


Blinding mists are common, and, moreover, sometimes come on almost without warning. From its conformation the land is full of streams. There are stones everywhere. What I found, therefore, as had others before me, was that as a consequence of the conditions to which I have referred, directions had been indicated by rows of stones for quite other than ceremonial purposes. Here, then, was a possible third origin. It was a matter of great importance to discriminate most carefully between these alignments, and to endeavour to sort them out. My special inquiry, of course, was to see if they, like their apparent equivalents in Brittany, could have had an astronomical origin. The first thing to do, then, was to see which might have been erected for worship or which for practical purposes.

In doing this there is no difficulty in dealing with extremes. Thus one notable line of large flat stones has been claimed by Messrs. R. N. Worth and 1 Continued from p. 34.

[ocr errors]

R. Burnard as a portion of the Great Fosseway (Rowe's "Perambulation," third edition, p. 63); it has been traced for eighteen miles from beyond Hameldon nearly to Tavistock, the stones being about 2 feet thick and the road 10 feet wide.

There are two notable avenues of upright stones at Merrivale; they are in close connection with a circle, and could have had no practical use. These stones, then, we may claim as representing the opposite extreme of the Fosseway and as suggesting an astronomical, as opposed to a practical, use; the adjacent circle, of course, greatly strengthens this view.

It is between these extremes that difficulties may arise, but the verdict can, in a great many cases at all events, be settled without any very great hesitation, especially where practical or astronomical uselessness can be established. But even here care is necessary, as I shall show.

The stones now in question, originally upright, are variously called avenues, rows, alignments or


Photo, by Lady Lockyer.

FIG. 17.-The Southern Avenue at Merrivale, looking East.

parallelithons. Their study dates from 1827, when Rowe and Colonel Hamilton Smith examined those at Merrivale (Rowe, op. cit., p. 31). Their number has increased with every careful study of any part of the moor, and doubtless many are still unmapped. The late Mr. R. N. Worth, of Plymouth, and his son, Mr. H. Worth, have given great attention to these monuments, and the former communicated a paper on them to the Devonshire Association for the Advancement of Science in 1892 (Trans., xxv. pp. 387-417).

A word of caution must be said before I proceed. We must not take for granted that the stone-rows are now as they left the hands of the builders. The disastrous carelessness of the Government in the matter of our national antiquities is, I am locally informed, admirably imitated by the Devonshire County and other lesser councils, and, indeed, by anybody who has a road to mend or a wall to build. On this account, any of the rows may once have been much longer and with an obvious practical use; and

1 Only yesterday (June 15) that excellent guide of the Chagford part if the moor, Mr. S. Perrott, showed me an avenue (Azimuth N. 20° E. true) near Hurston Ridge which is not shown in the 1-inch map.

those which now appear to be far removed from circles may once have been used for sacred processions at shrines which have disappeared.

Again, the rows of stones we are now considering must not be confounded with the "track lines" or boundary banks" which are so numerous on Dartmoor and are represented in Wiltshire according to Sir R. C. Hoare; these serve for bounds and pathways, and for connecting and enclosing fields or houses.

Dealing, then, with stone rows or avenues, which may be single, double, or multiple; any which are very long and crooked, following several directions, are certainly not astronomical; and it is easy to see in some cases that they might have been useful guides at night or in mist in difficult country with streams to cross. This possible utility must not be judged wholly by the present conformation of the ground or the present beds of streams.

For multiple avenues it is hard to find practical uses such as the above, and we know how such avenues were used in Brittany for sun worship. Mr. Baring Gould considers there were eight rows in an avenue on Challacombe Down 528 feet long; of these only three rows remain, the others being represented by single stones here and there (Rowe, p. 33). I shall have something to say about this avenue further


Although, as I have said, long rows bending in various directions are not likely to have had an astronomical origin, it must not be assumed that all astronomical avenues must be exactly straight. This, of course, would be true for level ground, but if the avenue has to pass over ridges and furrows, the varying height of the horizon must be reckoned with, and therefore the azimuth of the avenue at any point along it.

I think it possible that in the Staldon Moor row we have the mixture of religious and practical intention at which I have before hinted. Both Mr. Lukis and Mr. Hansford Worth have studied this monument, which is two miles and a quarter long. There is a circle at the south end about 60 feet in diameter, while at its northern end there is a cairn.

Where the line starts from the circle the direction of the row is parallel to many sight-lines in Cornwall, and Arcturus would rise in the azimuth indicated. But this direction is afterwards given up for one which leads towards an important collection of hut circles, and it crosses the Erme, no doubt at the most convenient spot. More to the north it crosses another stream and the bog of Red Lake. All this is surely practical enough, although the way indicated might have been followed by the priests of the hut circles to the stone circle to prepare the morning sacrifice and go through the ritual.

But there is still another method of discrimination. If any of these avenues were used at all for purposes of worship, their azimuths should agree with those already found in connection with circles in other parts of Britain, for we need not postulate a special race with a special cult limited to Dartmoor; and in my inquiries what I have to do is to consider the general question of orientation wherever traces of it can be found. The more the evidences coincide the better it is for the argument, while variations afford valuable tests.

Now, speaking very generally (I have not yet compared all my numerous notes), in Cornwall the chief lignments from the circles there are with azimuths 10-20° E. watching the rise of Arcturus, N. 24-28 E. watching the rise of the May sun, N. 75-82 E. watching the rise of the Pleiades. The

variation in the azimuths is largely due to the different heights of the horizon towards which the sight-lines are directed.

The conclusion I have come to is that these alignments, depending upon circles and menhirs in Cornwall, are all well represented on Dartmoor associated with the avenues; and further, so far as I have learned at present, in the case of the avenues connected with circles, there are not many alignments I have not met with in connection with circles in Cornwall and elsewhere.

This is not only a prima facie argument in favour of the astronomical use underlying the structures, but it is against the burial theory, for certainly there must have been burials in Cornwall.

In order, therefore, to proceed with the utmost caution, I limit myself in the first instance to the above azimuths, and will begin by applying a test which should be a rigid one.

If the avenues on Dartmoor had to deal with the same practices and cults as did the circles in Cornwall, they ought to prove themselves to have been in use at about the same time, and from this point of view the investigation of the avenues becomes of very great importance, because of the destruction of circles and menhirs which has been going on, and is still going on, on Dartmoor. We have circles without menhirs and menhirs without circles, so that the azimuths of the avenues alone remain to give us any chance of dating the monuments if they were used in connection with sun worship. The case is far different in Cornwall, where both circles and menhirs have in many cases been spared.

On Dartmoor, where in some cases the menhirs still remain, they have been annexed as crosses or perhaps as boundary stones, and squared and initialed; hence the Ordnance surveyors have been misled, and they are not shown as ancient stones on the map. In some cases the azimuth of the stones suggests that this has been the sequence of events.

It will be seen from the above that I have not tackled a question full of pitfalls without due caution, and this care was all the more necessary as the avenues have for long been the meeting ground of the friends and foes of what Rowe calls "Druidical speculations "; even yet the war rages, and my writing and Lieut. Devoir's observing touching the similar but grander avenues of Brittany have so far been all in vain; chiefly, I think, because no discrimination has been considered possible between different uses of avenues, and because the statements made by archæologists as to their direction have been quite useless to anybody in consequence of their vagueness, and last of all because the recent work on the Brittany remains is little known.

I began my acquaintance with the Dartmoor monuments by visiting Merrivale, and the result of my inquiries there left absolutely no doubt whatever on my mind. I was armed, thanks to the kindness of Colonel Johnston, the director of the Ordnance Survey, with the 25-inch map, while Mr. Hansford Worth had been so good as to send me one showing his special survey.

The Merrivale avenues (lat. 50° 33′ 15′′) are composed of two double rows, roughly with the azimuth N. 82° E.; the northern row is shorter than the other. Rowe, in his original description (1830), makes the northern 1143 feet long; they are not quite parallel, and the southern row has a distinct "kink" or change of direction in it at about the centre. The stones are mostly 2 or 3 feet high, and in each row they are about 3 feet apart; the distance between the rows is about 80 feet.

« PreviousContinue »