F site 
oO 
its anibera, each of which confilts of four cel/s, commu- 
nicating with one common tube, the excretory duct of 
the pollen. In the conitruétion of this name we run 
counter indeed to a precept of Linnzeus (Crit. Bot. p. 44), 
and we do fo becaufe in that inftance we think him in 
the wrong. After objecting, with reafon, to generic 
names too fimilar in found to each other, he is fome- 
what unmerciful in itigmatizing almoit all that have 
any fyllables in common, and wonders at Vaillant for 
ufing the termination 7/eca at all. The word furely in 
itfelf is unexceptionable ; and as all the generic names 
of Vaillant conitructed with it, even Tetragonotheca 
(which Linneus at firit retained), are now laid afide, and 
therefore there can be no ambiguity, we hope to be 
excufed for adopting ¢eca, as it fo precifely fuits our 
purpofe. 
Tetratheca probably belongs to M. de Juffieu’s order 
of Erica, not indeed that it anfwers well to his chara¢ters 
of that order, but it is allied to fome of its genera, efpe- 
cially Pyro/a. All its fpecies are {mall fhrubs with red 
flowers (varying to white), which retain their colour 
when dried. 
Tetratheca juncea has a {mall woody root, which has 
fome appearance of that of an annual plant. The /em 
is much branched, even from the bafe; the branches 
alternate, long and {lender, very acutely triangular, and 
almoit winged. Leaves mottly {mall and not numerous, 
alternate, lanceolate, entire. Szipude none. Each branch 
produces 
