TUNICATA.

PYCNOCLAVELLA AURILUCENS, Garstang.

This is a very beautiful little Compound Ascidian, a near ally of Clavelina, that I described in the journal of the Association in 1891. (Vol. ii. p. 53.) It has not yet been discovered in any other locality except the immediate neighbourhood of Plymouth, where two varieties are to be met with. Even at Plymouth, however, it is by no means common. The colonies are small in size, and consist of a number of pedunculated zooids which arise independently from a common encrusting base. The apertures of the zooids are plain and circular, not divided into lobes. The colour of the test is greenish, and each zooid is marked by a conspicuous streak along the line of the endostyle, either of a bright golden colour or opaque white. It inhabits very rocky ground, and is usually attached to irregular masses of Polyzoa and Annelid tubes, or may form a thin carpet on the stems of Delesseria. ARCHIDISTOMA AGGREGATUM, Garstang.

I first obtained this very interesting Compound Ascidian in the summer of last year, when dredging in fifteen fathoms, two miles south of the Mewstone, and have since procured it in considerable abundance in Plymouth Sound. It was briefly described in the Annals of Natural History for October, 1891, and in the Zoologischer Anzeiger, No. 378, for the same year. It resembles the last species in having a common encrusting base, from which arise club-shaped zooids. The zooids are not infrequently quite independent of one another; but usually they are aggregated together so as to form little clumps of zooids, partially joined together by a common test. Each zooid possesses two distinct apertures, which are always six-lobed. The colonies are found encrusting shells and stones, and are usually rendered very inconspicuous by the adhesion of sand-grains. It is the most primitive of the true Distomida, and with Pycnoclavella almost bridges the gap between this family and the Clavelinidæ.

CEPHALOCHORDA.

AMPHIOXUS LANCEOLATUS, Yarrell.

The Lancelet has been dredged off the Mewstone on a good many different occasions, and in its pelagic stage has been taken in the tow-net off the Eddystone, and close outside the Breakwater.

THE STONE ROWS OF DARTMOOR.

BY R. N. WORTH, F.G.S.

(Read at Plymouth, July, 1892.)

INTRODUCTION.

THE object of this paper is to reconsider in brief the twin questions of the origin and purpose of the most mysterious of the pre-historic monuments of Dartmoor, the rows of stones commonly known as avenues, parallelitha, or alignmentsnames we shall do well to discard in favour of the simple Saxon phrase stone rows, by which we first find them described in ancient documents.

It has been the misfortune of Dartmoor archæology to suffer from over-individualization. Thus Grimspound has been treated as exceptional in character because of its exceptionally perfect state, whereas it is but one enclosure of kindred type out of many. And so the special attention paid to the stone rows near Merivale Bridge is mainly responsible for the free use of the terms "avenues" and "parallelitha." Had it been clearly seen that while, on the one hand, there were many instances of such structures consisting of single rows, so, on the other, the number of rows at times increased up to two and twenty, the inapplicability of the word "avenues" would have been at once apparent. The essential feature is the row, not the number of rows; for although two are very commonly found, ones are as numerous on Dartmoor, and twos are really in a minority when compared with the general aggregate. So with the term "parallelitha"; it is correct enough up to a certain point; but there are partially double-membered rows, which never have been more complete; and there are sets of rows in which, so far from parallelism, there is marked and intentional divergence. "Alignments" is better than its companions; but here again the idea of straightness is imported,

whereas some of these rows, as already stated, are divergent, while others are more irregularly curved. The simple word "row," however, has the merit of expressing neither less nor more than the actual facts, and it can be defined without delimitation to express precisely the peculiarities of each individual case. Moreover, it is uncontaminated by hasty speculation or unwise theory. I think it preferable also to the final alternative "line," since it does not include the idea of absolute continuity.

Our special interest as an Association in these "rows" lies in the fact that individual instances are more numerous on Dartmoor than in any other part of the world. Twenty-seven Dartmoor examples are noted in this paper: eleven single; eleven double; one partially doubled; one double continued as a single; one group of three made up of a single and double side by side; one of three; and one of seven. Mr. Fergusson, in his Rude Stone Monuments, summed up all that was known up to 1872, and while there have been further discoveries since, I do not think that they have been in different fields. We get these rows in Devon (but not in Cornwall 1), in Somerset, Wilts, Kent, Cumberland, Caithness, Carnac and elsewhere in Brittany, one example in Lombardy, and doubtful ascriptions in Spain and Denmark. They are not recorded for Wales or Ireland, or for any other part of the Continent, though reported from Persia and India. Hence, all things considered, these monuments may be studied most fully on Dartmoor.

DARTMOOR.

We commence by a statement of the facts. I purpose first to give a description of the "stone rows" of Dartmoor and of their kindred elsewhere. The more important and exceptional of our local examples are dealt with most fully; and several of these have hitherto found no record, or have been very inadequately and incorrectly treated. The responsibility of the statements now made, where no authority is quoted, rests with the writer. Description given, we may proceed to draw some definite general conclusions, and pave the way for the solution of the problem, even should it be thought that solution cannot yet be reached.

And I should premise that the destruction of some of the

avenue save

"Remains of avenues' are indeed said to have existed near Kilmarth, but the Rev. Mr. Lukis could find nothing approaching an the eight stones called the "Nine Maidens," near St. Columb. Moreover, these average 68 feet apart, and bear no likeness to the true rows."

2

stone rows of Dartmoor is certain. Those which continue are the remainder of a more numerous-though probably never very numerous-body; nor is it likely that all extant are as yet recorded. But let inquirers beware of wrong identifications. Some of the old track stones which markedindeed, still mark-routes across the Moor have ere now been mistaken for these structures; so have mere field boundaries; and the facing-stones of ancient hedges whence the earthen filling has disappeared have been dubbed "parallelitha." But the true pre-historic stone row is sui generis. All the rows in the southern and western quarter of Dartmoor have been specially visited for the purpose of this paper; and I have seen nearly all the remainder. They are found chiefly in the valleys of the Teign, Avon, Erme, Yealm, Plym, Meavy, and Walkham; only casually in the watershed of the Dart; they are as yet unreported from the Tavy and the Okements and the Taw, though there is one notable example near the latter river; and this particular distribution has, without doubt, its meaning.

Drewsteignton.

A very complete system of rows and circles is said to have existed in close contiguity to the Drewsteignton Cromlech. Unfortunately they cannot now be traced; and one may be pardoned a little scepticism as to whether their remains ever fell so exactly and elaborately into place as is shewn on a plan prepared by the Rev. Wm. Grey in 1838, and reproduced by the late Mr. Ormerod, F.G.S., in his Notice of the Fall and Restoration of the Drewsteignton Cromlech, as

2 Since this paper was read Mr. R. Burnard has informed me of the exist ence of a hitherto unrecorded "row" at Assycombe Hill; and has kindly supplied me with the following extract from his field notes. Mr. Burnard has a scale plan of this row in preparation :

"Stone Row, on western slope of Assycombe Hill, which lies N.W. of Warren House Inn, distant about 14 miles from the latter. (See Sheet 69, N.E., 6" new 'Survey of Dartmoor.') The row, which starts from a ruined cairn, consists of 84 standing stones, and extends 800 feet E. and W., running in two rows down the hill towards Assycombe. It terminates towards the W. in a stone lying prone, 5′ 9′′ long. 23' wide. The avenue is from 5' to 6' wide, inner measurement. The stones composing the row are 18" high, and down to 6"."

Mr. Burnard, in his presidential address to the members of the Plymouth Institution, called attention to the existence of remains at May's Newtake, in the Swincombe valley, which may possibly represent a dilapidated "row." Three small menhirs stand in a line about two feet apart from each othertwo 4 feet high, and one 3 feet. At an interval of some two hundred yards there are two or three stones in a straight line with them, which sug gested that the menhirs "might be portions of a previously existing stone avenue." I agree with Mr. Burnard in thinking the evidence uncertain; but the fact should clearly be mentioned in this connection.

reprinted in 1876. Mr. Grey stated that 110 yards to the west of the cromlech he found

"two concentric circles of stones, the inner circle having entrances facing the cardinal points, that to the north being 65 paces in length and 5 broad. The outer circle, besides these, has avenues diverging towards north-east, south-east, south-west, and north-west. A smaller circle seems to intersect the larger, of which the avenue eastwards is very evident."

These remarks, entered by Mr. Grey in his journal at the time, unquestionably represent the impression made upon his mind, and the map as reproduced by Mr. Ormerod was approved by him.

Nevertheless I cannot but feel very doubtful-not of the facts observed, but of their interpretation. A double row leading towards the cromlech is shewn on the plan clearly enough; so is another at right angles to it; the outer circle is also distinctly set forth. But apart from the stones of the "avenues," the inner circle is represented by two members only; and the materials indicated for the diverging "avenues" and intersecting circle are quite insufficient for such positive deductions; while there are patent irregularities which it is impossible to bring into the scheme propounded. Nothing short of an exact professional survey could supply an adequate basis for the acceptance of Mr. Grey's plan in its entirety; and that has long been an impossibility, for when Mr. Ormerod examined the ground in 1872 he did not find a single stone which could be identified on Mr. Grey's map (the distances on which, by the way, were given only in paces), though others in the vicinity (not on the map) were observed! Either they had all been removed in the interim, and others provided, or the plan was excessively inaccurate. Mr. Ormerod, moreover, with all his inquiries, was unable to find more than two people who recollected the remains at all; and they remembered an "avenue" leading east and west in a line with the cromlech, but nothing more. The only other references to these remains are by Polwhele, who says: "At Drewsteignton the Cromlegh is placed on an elevated spot overlooking a sacred way and two rows of pillars, and several columnar circles."3 While elsewhere, in describing the

3 History of Devon, 154; Historical Views of Devonshire, 94.

History, i. 150; Historical Views, 61. The full passage is: "Towards the west of the Cromlech I remarked several conical pillars, about four feet high. On the north side there are three standing in a direct line from east to west. The distance from the more western to the middle was two hundred and twelve paces—from the middle to that on the east, one hundred and six