6 
into natural groups, as well as for the critical comparison of the 
numerous descriptions already published by different authors, and 
Jovellana, wdiich I only know from the works of Cavanilles and 
Ruiz and Pavon, appear to me to connect Alonsoa with Scrophu- 
laria, and perhaps with them also should be associated the Bcea of 
Commerson. But on account of the sihquose fruit I have some 
doubts of the latter being a true Scrophularinea. 
Scrophularia has been worked up with great care and detail by 
Wydler, to w^hose w’ork I have only to add the Indian species men- 
tioned below. 
The second tribe, the Hemimerideae, form a small group of four 
genera, differing from each other in the dehiscence and form of the 
capsule, and forming the passage from Verbasceae to Antirrhineae, 
allied to the former by the explanate short-tubed corolla and sub- 
unilocular stamina, to the latter by the sack or spur at the base 
of the corolla. The connection of the two first genera, Angelonia 
and Thylacantha*, both American, with the Cape Hemimerides is 
not usually recognised, but appears to me to be intimate. The 
species of the latter genus and of Nemesia are probably numerous, 
and stand much in need of the labours of some better Phytographist 
than Thunberg. 
The tribe of Antirrhineae are better known than any 'of the 
order, thanks to the elaborate work of Chavannes, to which later 
discoveries have but little to add. The question, whether Jussieu’s 
Galvesia is to retain that name, or to exchange it for Agassizia, and 
whether Lophospermum and Rhodochiton be but one, or whether 
they are two distinct but adjoining genera, are of little importance, 
and can only be settled in a general work by a Botanist of authority. 
Under Salpiglossideae I have brought seven genera, but with 
some hesitation as to some of them. Brunsfelsia has a baccate 
fruit (which I have not had an opportunity of examining), and 
may perhaps be a true Solanea. Anthocercis is probably here in its 
right place, but I find I have not been quite correct in the distinc- 
tive character corollcB limbo subrecto, which does not apply to 
Labillardiere’s original species, it might be corrected to limbi lobis 
acutis. 
Lyncea (Schlecht in Linnaea, 5. 108.) w^hich I had overlooked, 
is apparently near Brunsfelsia, but not having myself seen the plant, 
and the fruit not having been yet observed, it is impossible to deter- 
mine w'hether it even really belongs to this order. 
Schizanthus, in many respects so very like Salpiglossis (especially 
when past flower), appeared to me to connect this genus with Col- 
linsia, and the latter genus to be placed between Schizanthus and 
the Antirrhineae ; but ulterior observations have pointed out to me a 
f By an error of the press, this genus is called Phylacanthus in the above-men- 
tioned Enumeration. 
