1896.] Pra:EGp:r. — Botanical Subdivision of Ireland. 35 



the Midland Great Western Railway from Oranmore, at the head of 

 Galway Ray, to Ballinasloe on the River Suck, the eastern boundary of 

 the county. It may Ije remarked here that the Aran Islands, though 

 part of Co. Galway, belong botanically to Co. Clare, and are so treated 

 in Cybde Ilibcrnica ; and that Inishbofin, formerly included in Co. Mayo, 

 is now a part of West Galway, to which it naturally belongs. 



DoNEGAiy. — This large county (1,870 square miles) should evidently 

 form two vice-counties, in order to keep the variation of size of our 

 ultimate divisions within reasonable limits, and thus ensure that a 

 statement of the number of county-divisions in which a plant occurs in 

 the country may be a tolerably correct indication of its area of distri- 

 bution. 



The boundary which I suggest is the roughly east and west line which 

 separates the baronies of Inishowen and Kilmacrenan on the north 

 from Raphoe and Bo^dagh on the south. This line crosses the Inishowen 

 isthmus at its narrowest point, follows the shore of Lough Swilly, and 

 then the River Swilly almost to its source, and descends to the western 

 ocean along the course of the Gweedore River ; and it divides the county 

 into two almost equal parts. 



The whole of Ireland, 32,513 square miles, is thus divided 

 into 40 portions of as nearly equal size as conditions will per- 

 mit, the average area of these portions being 813 square miles. 

 This size is almost identical with the average size of Watson's 

 112 vice-counties of Great Britain, which is 804 square miles. 



Next, as to the order in which the counties and vice-counties 

 should be numbered. Watson numbered the British provinces 

 I. to XVIII., commencing with S.W. England and ending 

 with the extreme north of Scotland. The vice-counties he 

 numbered in the same order, those included in Province I. 

 being numbered i to 6, those of Province II. 7 to 14, and so on. 

 Babington proposed a similar method for Ireland, but the 

 result is not satisfactory^ The Irish " provinces " are not 

 numbered regularly from south to north, but the numbering 

 runs first up the east coast, and then drops back into the 

 south-west ; and this absence of regular progression becomes 

 accentuated if the vice- counties are numbered in the sequence 

 of the provinces ; when, for instance, we suddently pass from 

 lyouth (127) 120 miles south-westward to Limerick (128). 

 It will be generally admitted that the best scheme, and the 

 most natural, is one which will show a regular progression 

 from south to north — from a higher temperature to a lower : 

 with such a system, the largeness or smallness of the numbers 

 in the list showing the county-distribution of a species, will 



