390 
44. RUBIACEiE. 
Now this pi. is plainly, both by the syn. and diagnosis, G. 
eUipticum Willd., DC. &c., and to it he appends, unfortunately 
(p. 157) as a var. /3, his former G. rotundifolium of ed. 1. 
p. 108. 
The name rotundifolium belongs therefore plainly and simply 
to the pi. originally so designated by him in ed. 1. And his 
G. rotundifolium ed. 2. p. 150 (excl. var. /3 p. 157) merges into 
a syn. of G. ellipticum Willd., DC. &c. 
It can scarcely be considered otherwise than a mere accident 
or inadvertence that in his Herbarium he has inscribed propria 
manu the name rotundifolium on an indubitable specimen of a 
pi. subsequently called by him (Mantissa, p. 38) Aspeimla Ice- 
vifjata and at variance in toto with the syn. and diagnosis of 
both his G. rotundifolia in Sp. PI. ed. 1 and 2. 
The widely diffused European true G. rotundifolium L. 
(ed. 1) is quite distinct from every form of the Mad. and Can. 
pi. by its altogether smaller size and more delicate habit, its 
slender simple erect herbaceous st. scarcely above 6 or 8 in. high, 
its smaller pedunculate leafless lax few-fld. less branched or 
decompound panicle, with the subdivisions longer and more 
spreading and the fl. more remote. Its correct synonymy will 
stand thus : 
G. rotundifolium Linn. Sp. PI. ed. 1. p. 108 (not Herb. Linn. !). 
G. rotundifolium (3 Linn. Sp. PI. ed. 2. p. 157. G. rotundi- 
folium DC. Prodr. iv. 599 (excl. syn. Linn., Bocc. et Moris.). 
G. rotundifolium Koch ed. 2. p. 364 (excl. syn. Linn.). G. ro- 
tundifolium Gren. et Godr. ii. 17 (excl. syn. Linn ). G. rotun- 
difolium Lam. Diet. ii. 577, 578 (excl. syn. Bocc., Moris., 
Barrel, et Linn.). Asperula Iceviyata 8 Lam. Diet. i. 298 (excl. 
syn. Moris., Barrel, et Boc.). — Icon., Moris, t. 21. f. 4 ; Barrel, 
t. 323. 
1 have lately received as “ sp. Mad. nov. ?” from my vigilant 
and unwearied friend the Barao do Gastello de Paiva, a single 
dried example of a pi. found recently by one of his collectors 
on u a rock above the Icehouse ” in Mad. It is entirely smooth 
or glabrous, dwarfish with small 1. (2-4 lines long, 1-2 broad) 
and depauperated few-fld. cymes exactly as in Moris, t. 21. f. 4 
and Barrel, t. 323, i.e. G. rotundifolium L. (ed. 1) verum ; in- 
somuch that I was inclined at first to think it really was that 
