368 IX. ORDINAL SUMMARIES. 



two or more orders, the sequence of tbeir names is then determined by 

 the figures set in the third column. The second column of figures shows 

 the per-centage of the several orders in the whole flora ; thus, Composit<B 

 count nearly to nine and a half in the hundred species ; Umbelliferce 

 make up four in the hundred ; Geraniacece, and all others placed below 

 that order, being only a fraction per cent. 



The third column of figures repeats the census of the orders, as before 

 introduced into the ' Summary,' and explained on pages 231 and 232. 

 It is the sum total of the comital (always including the vice-comital) 

 numbers placed immediately before the several names in the ' Census of 

 Species,' pages 234 to 271, re-arranged into orders. The total of 

 counties for each order being divided by the number of its species, an 

 average of frequency or diffusion is ascertained for those of each order, 

 as shown by the fourth column of figures. But it will readily be under- 

 stood, that the species become too few in the lower portion of the series, 

 to yield averages suitable for comparison with those of orders much 

 higher in the series. And in comparing the averages one with another, 

 due allowance must be made for any artificial peculiarity of the order 

 which may alter its proper average ; for instance, that of Rosacea is 

 doubtless reduced too low by the numerous segregate species o{ Rubus, 

 the comital distribution of which is less fully ascertained, than that of 

 other more true or more understood species. Doubtless, also, the 

 segregate species of Hieracium, many of them recorded from very few 

 counties, somewhat reduce the average for the Composite. Although 

 liable thus to some irregularities, the comital census and averages serve 

 usefully to correct false ideas which might otherwise be suggested by 

 the number of species in an order, if alone regarded as a measure of its 

 predominance in the vegetation of the island. For example, the orders 

 Filices and Orchidacem may be cited as on equality in the numbers of 

 their species ; but on looking to the comital census or average frequency 

 of the species, we find the ferns thus proved to prevail much more than 

 the orchids. The comital average for the order Filices is 46, or five 

 above the general average of 41 ; that for the order Orckidacete being 

 only 32, or nine below the general average. These two orders are very 

 suitable for the comparative illustration ; having no introduced species 

 included with them, and both being botanical favourites, much looked 

 for, and their localities usually recorded. 



The Jiflh and sixth columns are introduced for comparison with the 

 first. They show the numbers of species for the same orders in Europe 



