— 334 — 



surface of the leaf to be paler than the upper, or ventral, except 

 for interrupted narrow lines of dark green. The name A. muUilineafa 

 has been used to designate one such case in the common "Lechu- 

 guilla", where, however, the often nearly white under surface assumes 

 a much deeper tone under certain conditions of growth even in 

 the same individual. Rarely, the Littaeas are mottled with whitish, 

 purplish or bronzed blotches on one or both surfaces (Ä. Eovel- 

 liana or Terraccianoi), a State that is quite common and sometimes 

 characteristic in the annual-leaved facemose species which have 

 been properly segregated under the generic name Man/reihi. 



Typ es of Variegation. 



None of these normal markings are included in the following 

 consideration of variegations, which are abnormalities, comprising, 

 in these genera, brilliant types of contrasting coloration though 

 aside from green this is limited in nature leaves to white and 

 yellow, a shade of rose color being sometimes superadded in 

 young leaves. 



Essentially, the variegated forms are of two types, (l) with 

 the margin white or yellow, and (2) with the margin of the normal 

 green. hi both, the marginal coloration is variably more or less 

 decurrent into the central part in lines or bands of different width, 

 and in the yellow-centered forms isolated lines of bright or pale 

 green — the latter best seen when the leaf is held to the light — 

 are not infrequent. When the intrusions of green are long and 

 numerous, and equal or exceed the white or yellow in width, they 

 form of the median-banded a seeming third form, that of striation. 

 The limits between the contrasting colors are usually very sharp, 

 though it sometimes happens that a bright green margin or stripe 

 is flanked by a zone of paler green. One Linnaean species, Agave 

 americana, epitomizes these types. 



Classification and N o m e n c 1 a t u r e. 



Horticultural nomenclature is attended with difficulties peculiar 

 to itself. Botanists have never given much heed to suggestions such 

 as Bossin's'), that the more fixed forms receive binomial or 

 trinomial Latin designation quite apart from that of their known 

 or supposed wild prototypes; but the more comprehensive and 



') Bull. Congr. Intern, de l!ot. et d'Hort. Amsterdam. 352 (1866). 



