150 THE JOURNAL OF BOTANY 
which, when fully developed, is 9-10 mm. in length in all of the 
specimens. ere ee minor divergencies, however, in other 
respects, the calyx of S. Marquandit being much lighter in colour, 
suffused generally with a pinkish purple, with the upper lip a 
little narrower and a trifle recurved upwards at the apex after the 
fashion of S. pratensis. The glands of the upper lip, also, are 
fewer and on longer stalks, and are freely interspersed with much 
longer pilose hairs than are met with on the upper side of the 
calyx in the British plants. 
‘The comparison of the corolla of S. Marquandii with that a 
the Hampton n Court form proved most interesting. In Mr. Mar 
quand’s specimens, which are somes this organ was uni- 
formly utes ndrous, from 12 to nearly 15 mm. in total length, 
with the upper lip galeate in wapiile and the lower spreading and 
deflexed. e smaller flowers were indistinguishable in shape 
ft 
greater contraction of the throat within the calyx; and the chief 
difference in colour lay in the corolla of the Guernsey plant being 
wholly concolorous, without the oes spots at the throat, and its 
hue somewhat brighter and more tinged with lilac Mr. Mar- 
quand’s largest Sowers. showed a slightly ae upper lip, and 
gen approximated to the large corollas of the Lizard ‘lass 
of 190 
ut this type of corolla is not invariably produced in i Mar- 
f Ww 
q , for one of Mr. Bennett’s specimens . Reeves, 
pete 1886) ao smaller and differently shaped pg ae not 
exceeding 10 n the central spike; and in another (M. Dawber 
May, 1886) the I Seal branches secs nothing but small flowers, 
some of which are of semi-cleistogamous form, and only 7 mm. 
long. It is thus evident that this plant is subject to the same 
polymorphism of the corolla as other Verbenaca forms, although 
it produces a much greater igegreieyt of proterandrous flowers 
than we are accustomed to see in Brita: 
From these considerations it will o seen that this Salvia 
cannot be named S. horminoides, but agrees essentially in its 
slender habit and narrow leaves with the restricted S. Verbenaca, 
t. ” and eit the foes and aye of the same plant, 
Gallitrichum pallidiflorum, in Jordan & Fourreau’s Icones, both 
the specimen and the figure showing not only the same habit, 
foliage, and corolla-colouring, but even the same peculiarities of 
the form, glands, and agrenk of the calyx. 
If this be admitted, it may be considered that of the forms 
shown in the poem clavi we have in Britain the subspecies 
. horminoides Pourr ected with its variety incisa Benth. by 
various presen a ae and usually producing cleistogamous o: 
