THE LATE REV. W. H. PURCHAS 81 
ie band of workers, he investigated and recorded energetically. 
The work flagged on his removal from the coun nty in 1857 and the 
mulliphesaan of ministerial and family cares; a in 1889 its 
results were at length published by him, in conjunction with my- 
self, in the Flora of Herefordshire. 
All Mr. Purchas’s botanical work was painstaking and thorough, 
spared sunae neither time nor trouble in the investigation of 
critical point He was self-distrustful to an excessive degree, and 
was very averse from sakislane results until he felt perfectly sure 
of his ground. This, while diminishing from the fruits of hig 
labours as regards science in general, r rendered his knowledge of 
He was early attracted by what are called Moe ie genera ’’— 
Rubus, Rosa, Hieracium, Epipactis, and others. These it was his 
wont to watch, to study, to go again and eas over the points 
demanding attention, and often to end habe he began with ‘ After 
all, I doubt.’’ The tangle of the fruticose Rubi was a pas mee 
gave full scope to his peculiarities of temperament. Here had 
the advantage, on the one hand, of living in a district ae os 
to be prolifie beyond any in Britain in bramble forms, and, on the 
other, of the friendship and co-operation of the band of early 
bramble students—Bloxam, Coleman, Newbould, and Babington. 
The last-named was a frequent visitor to Lingwood at Lyston, 
near Hereford, and Purcias joined in the investigations wah led 
to the publication, in 1870, of Babington’s British Rubi. 
In those days, and indeed until the great enlargement of o 
knowledge of the genus in the nineties, through the help of the te 
Mr. Archer Briggs, Prof. Focke, and Rev. W. M. Rogers, the study 
of British brambles was a hopelessly puzzling affair ; it involved 
trying to fit some 150 forms into 80-40 descriptions, and resembled 
the attempt to force 150 apples into a basket designed to contain 40: 
as fast as one was forced in another jumped out ; “and this confusion 
Was worse confounded by innumerable inconsistent iba aipaiat 
e eas 
Purchas’s name—Rubus Purchasianws—is a very local form, ai 
ordinarily plentiful near Ross ; the synonymy of this form is a 
good illustration of the fluctuation of these early discussions. 
It is characteristic of Mr. nico’ that, naan for many a 
this genus was his special subject of study, he never named mo 
than one Brit an ie and only (I believe) 6 ioe British 
plants (hawkweeds). Here again his real w ork is by no means 
adequately ponienentla by the plants which he Neaieibiad. He knew, 
more thoroughly than any other man, the weeds inhabiting the 
lower Herefordshire MP valley, and the site and North 
Goa OF ay ou. 42. [Marou, 1904. si 
