September 5, 1907] 



NA TURE 



479 



It is curious to note that the total of fairs in February, 

 May, August, and November, with pre-Lenten and Whitsun 

 fairs, is 1425, just ten less than the total of true May- 

 year fairs ^lus all the Church-year fairs which I would 

 claim for that year. 1 can, therefoi'e, add the Michaelmas 

 and St. Andrew fairs to the last total obtained and make 

 it 1571, or a round fifteen hundred May-year fairs, nearly 

 two-thirds of the total number of solar, as distinguished 

 from mere calei:dar, fairs. 



Wales makes but a very poor show in Owen's list. 

 Instead of five complete May-year series, I have noted 

 ten in North Wales alone ; but it serves the present pur- 

 pose to keep Wales in the background. 



It is satisfactory to find seventeen complete series. The 

 case of Weldon, Northamptonshire, is eloquently put as 

 follows : — " First Thursday in February, May, August, and 

 November." There are, however, ten other combinations 

 of May-year quarter days, each of which tells the same 

 tale, and there are 277 places where the May-year seasons 

 are observed by fairs, that is, where more than one May- 

 year fair is held. The relative prevalence of the eleven 

 combinations may be shown as follows : — 



Each May-year quarter day enters into seven combin- 

 ations, which reminds me of the story invented, I believe, 

 to account for the popular name of the parish from which 

 I write, Yr Hin Bluiyv, the Old Parish. The story goes 

 that a stone-cutter carved the figure 7777 on a grave- 

 stone intended to commemorate a man whose age was 

 twenty-eight. By the way, multiplying the May-year 

 sevens together would be a good way to remember the 

 number of solar fairs we have made out, 2303 for 2363, 

 leaving a margin of sixty for possible errors in such a 

 large estimate. 



No combination of figures affects the supremacy of May 

 Day. February enters into combination with other 

 quarters at 80 places, May at 230, August at 139, and 



NO. 1975, "^'OL. 76] 



November at 183. Generally, the ratio seems to be : — 

 February = I, May = 3, August = 2, November = 2. 



But in the two most decisive factors there is not much 

 to choose between May and November. These two factors 

 are the prominence of the summer half of the May year 

 and the fact that the astronomical dates are still observed 

 at ninety-six places in remarkably even numbers — May, 55 ; 

 November, 41. May 6 is associated with St. John the 

 Evangelist. Such association may have helped to preserve 

 the date ; but no such Church sanction, favour, or sup- 

 port has been given to November 8. There are four 

 places in England where both dates are still observed. 

 In several instances where the astronomical dates have 

 been preserved I note a startling parallelism between the 

 dates and the prominence acquired by those places in 

 tradition and archaeology. 



List of Places where Fairs are held on May 6 and 

 November 8. 



