\ f > 
334 39. CRASST7LACEJE. 
versal Mad. pi. It is true that Solander’s restriction of his “ S. 
glandulosum ” to a single Mad. locality, u Paul do Mar in muris 
— affinis S. canariensi ” (MS. list of Mad. pi.), is difficult to be 
explained : but it is still harder to suppose that he or Masson 
could have overlooked a pi. so common, striking and peculiar 
as the present sp,, for which however S. glandulosum is the 
only name (except indeed the still more obscure “ paniculatum ”) 
remaining unappropriated on his list of Mad. Semperviva. And 
although DeCandolle’s very restrictive comparison of S. ciliatum 
Willd. with S. glandulosum (Mem. ii. sur les Crassul. p. 59) 
seems quite opposed to the identity of the latter (sensu Candol- 
leano ) with my pi., and although, again, S. glandulosum Haw. 
Rev. 65 (not Syn. 166) with “ fimbriated ” hypog. glands was 
also probably something different, — still the original sp. of Aiton 
was, I can doubt no longer, founded principally if not solely 
on the common Mad. pi. to which I have assigned it. 
S. tabulceforme Haw. Suppl. 69, Rev. 63, a strictly Can. pi., 
has been often confused with the present, having been first set 
down by Haworth and then by DC. (Prodr. iii. 412) erroneously 
as a Mad. sp. It is however easily distinguished by the re- 
markably long and regular pectinate close-set fringes of the 1. ; 
and it is indeed an altogether larger pi., differing no less by its 
more distinct sometimes (at least in cultivation) once- or twice- 
branched subperennial suffrutescent st. bearing often 2 or 3 ro- 
settes, than by its perfectly smooth 1. elegantly and regularly 
fringed with long close-set soft white cartilaginous simple cilia. 
Though entirely omitted by Webb, it occurs in vast profusion on 
the N.W. coast of Tenerife along the seacliff road from Realejo 
to S. Juan de laRambla, and also beyond Icod de los Vinos on 
the road to Garrachico, with a few pi. intermixed here and 
there of the true Mad. S. glandulosum, which has equally escaped 
the observation of Webb and other botanists as a Canarian pi. 
The principal indigenous Mad. Semperviva curiously corre- 
spond with several of the Can. sp. Thus the Mad. S. divari - 
catum, villosum, glutinosum find glandulosum represent both in 
habit and abundance the Can. S. jmnetatum DC., subvillosum 
Lowe, canariemc L. and tabulceforme Haw., all except S. glan- 
dulosum being strictly confined to their respective group of 
islands. 
