April 9, 1908J 



NATURE 



531 



E. ; 



alcing the stone and its supporters of N. 61° i- 

 the azimuth of the May sunrise with hills 2° hi_ 

 N 61° 30' E. 



The other exceptional case is at Garrol, where 

 there has been great disturbance, and where, as at 

 .\rdlair, the length of the recumbent stone lies in 

 the direction which points to the rise of the 

 Mav sun, the mean of many measures giving 

 N.'6i° 45' E. 



My measurements of the May-year circles were as 

 follows : — 



May-year. Sun's Declinatiou 16° 20' A". (May 6, 

 August 8). 



Some of the previous measures in Cornwall may 

 be given for comparison : — 



Monument at — 



Bo«cawen-un — 



Circle to two large men 

 hirs 



Merry Maidens- 

 Circle 10 Fongou 



Tre(;eseal — 



Circle to Longiton^ — 



Longston.;(Tregeseal)-- 

 To W. Lanyon Quoit .. 



Down Tor- 



Directijn of avenue 



May. 



St. Cleer - 



Holy well to Trev..-thy 

 cromlech 



Lesquoit crcmlech — 

 Orientation of cromlech. 



Druids Altar (Pawton)-- 

 Orientaiion of cromlech. 



May 6 Aug. 8 



May 

 May 

 May 4 Aug 



S Aug. 

 S Aug. 



In addition to these, I have found ' that Lukis ' 

 and Borlase ' give plans of a number of cromlechs in 

 Cornw-all which appear to be oriented to the May sun. 



They are as follows : — 



Cromlech. 



Authority. 



I-anyon Quoit Borlase; plate: 



.Mulfra Quoit Lukis : plate .\i: 



Chywoone Quoit Lukis; plate xx 



Zennor Quoit Lukis ; plate xx 



three Brothers Grusilh Lukis ; plate x.> 



Mean of .above 



N. 64' 1 



Assuming an elevation of the horizon between 3° 

 and 1°, this mean value is the exact azimuth of the 

 May sunrise in Cornwall. 



I next give details touching the solstitial circles. 



1 See -Nature, No. 1987, vol. 1 

 - ' The Prehistoric Sti 

 ^ " .\ntiquities of Coi 



With these, accurate measurement is a difficult matter, 

 and, as the determination of the date of erection from 

 the variation of the obliquity of the ecliptic entails 

 very precise measures, I content myself with pointing 

 out that thfe declinations are probably solstitial and 

 that they agree, in the mean, with the values pre- 

 viously obtained for the English solstitial circles. 



I append some measures made in the south of 

 England for comparison : — 



Stonehenge — 

 Direction of a 



Stanton Drew- 

 Great circle t 



. N.E. circle 



Bi 



Centre of circle to fine menhii 

 Tregeseal — 



Centre of circle to holed stone 

 Longstone (Tregeseal) — 



To Men 



l-Tol . 

 . of abo 



N. 49 34 E. 



23 54^ 

 23 49 



General Conclusions. 

 Should subsequent inquiries confirm the balance of 

 argument against the use of Capella, we shall be 

 led to the following conclusions: — 



(a) Dealing with the circles already measured by 

 me in the two localities, the Aberdeen circles are 

 more than a thousand years younger than those of 

 Cornwall and the west coast ; and here we have one 

 reason why the east-coast circles are dissimilar, and 

 those at Callernish and Stenness to the west are simi- 

 lar, to the Cornish circles. 



(b) With this great difference of time to deal with, 

 we have also probably a difference of origin between 

 the West coast and East coast swarms. 



(c) As the May year is still supreme in Scotland 

 generally, it is clear that the solstitial Aberdonians 

 were at some point of time overpowered in influence 

 by a return wave from the west of Scotland. 



Since my return from .Aberdeen Mr. Horton Bolitho, 

 one of the hon. secretaries of the Cornwall branch of 

 the Society for the .Vstronomical Study of Ancient 

 Monuments, and whose knowledge of the Cornish 

 alignments is second to none, has sent me the follow- 

 ing information concerning the circles in Perth- 

 shire : — 



" I examined six circles in Perthshire last year and 

 this, but I found no trace of recumbent stones such as 

 are associated with the .-Vberdeen circles. The Perth- 

 shire circles closely resemble the circles of Cornwall, 

 showing traces of May-year and solstitial alignments 

 with at least one clear use of a clock-star in azimuth 

 N. 18° E. Trees prevented any fine measurements 

 being taken, and local assistance in searching for out- 

 standing stones was lacking. In two of the best pre- 



1 .At Sunhoney, as the recumbent stone was curved and irregular, it was 

 simpler to measure directly across the circle at right-angles to the length of 

 the recumbent stone ; the magnetic azimuth thus obtained was 71' 20 . 



NO. 2006, VOL. 77] 



