AM A RYLLIDACEA2. 
103 
The variation in the form and colour of the flower of A. 
pnlchra, and the two coloured varieties, as well as of the seed- 
lings of A. Cummingiana from imported seed, should render 
botanists very cautious not to multiply species freely on the 
appearance of such diversities in natural specimens from 
different localities, and makes it very difficult to fix on the 
true distinguishing features. Little attention is to be paid to 
the length of the style in Alstrcemeria ; its maturity is very 
tardy ; it is very short at first, grows out slowly, and at last 
the stigma, which had appeared to be simple, expands and 
becomes trifid and patent. The anthers discharge their 
pollen long before the maturity of the style. The late 
developement of the stigma should make the genus very liable 
to spontaneous intermixtures of the species, but render it 
difficult to obtain artificial crosses. I failed in getting seed 
at all from an attempt to fertilize the red peregrina by the 
white, which must have arisen from having neglected the 
proper moment for fertilizing the style or for selecting the 
pollen. 
A. Psittacina, as well as hremantha and aurantiaca, 
flowers well in the open ground, if covered witli straw or a 
thick coat of leaves in the winter. The soil should be light, 
and the tubers set pretty deep ; and any heading that would 
throw the wet off’ in the winter will be found advantageous. 
It is absolutely necessary to pick the slugs oflP the border, 
which will otherwise devour every shoot at its first appearance 
above ground : and it will be found advantageous to cover 
the bed in the spring with dry sawdust which the slugs do 
not like to crawl over, and it will keep moisture in the 
ground. A top covering of peat is also disagreeable to slugs, 
which I find very troublesome in biting the flower-stalks of 
Gladioli on sandy loam, but they rarely do so on a border of 
black earth. 
16. Collania. — Stem rigid, erect, curved at the summit; 
leaves rigid (not reversed ?) umbel pendulous ; ger- 
men turbinate, the operculous base of the style be- 
coming enormously enlarged, and forming the main 
part of the fruit ; perianth rather tubeformed, and 
not at all patent. (Pericarp probably soft and pul- 
paceous, since the fruit in at least one species is sweet 
and eatable.) Filaments and style straight ? 
1. Involucrosa.— PI. 9. Specim. Matthews, Peru, 863. 
