452 NATURAL HISTORY OF PLANTS. 



sometimes compound. They are usually thick and persistent, 1 rarely 

 caducous, replaced in Cassytlia by small scales inserted on the para- 

 sitic filiform stems, which cling by suckers to the neighbouring 

 plants. The flowers are sometimes in simple spikes or racemes, but 

 far more frequently in cymes or ramified racemes of cymes. The 

 floral receptacle varies greatly in depth ; it is rarely convex, oftener 

 flat or concave, very frequently hollowed into a deep sac or pouch 

 bearing the perianth and androceum on its rim. This pouch is some- 

 times accrescent and persistent at the base of or around the fruit, 

 which it may even entirely envelope ; sometimes it separates earlier 

 or later from the pedicel, either by its base or some way up, bringing 

 the perianth away with it. The indusium formed by it around 

 the fruit will thus vary greatly in height, no less than in consistency; 

 it is usually dry, but sometimes fleshy as in Cassytha. The type of 

 the flower is usually sf ; but we may find v / ', %f 9 or \/ occasionally. 

 The androceum consists of one or more whorls ; four alternating is 

 the usual number. Certain of the stamens are introrse, certain ex- 

 trorse ; some have lateral glands, while others lack them entirely : 

 the two or four valves by which they dehisce also varying from ex- 

 trorse to introrse. Some of the stamens may be sterile ; when all 

 abort the flowers may be diclinous. The form of the stigmatiferous 

 end of the style varies ; the floral pedicel, usually remaining cylin- 

 drical below the fruit, is sometimes dilated to a variable extent and 

 club shaped. These are the variable characters which have served 

 to distinguish the genera and the eight series of this order. We 

 proceed to give the general features on which our classification of 

 these last depends. 2 



I. Cinnamome.e. — Flowers usually hermaphrodite; staminal whorls 

 4; four stamens of the two outer fertile and introrse; of the third whorl, 



1 In several genera (Cinnamomum, DIespilo- have come into contact with any animal what- 



daphne, Ocotea, Phcebe, &c), axillary to the ever. But it is not impossible that the great 



secondary ribs, especially near the base of the development sometimes assumed by these cavities 



blade, we find more or less marked projections (as in Ocotea bullata, fattens, &c.) may be really 



above, corresponding with depressions or pores, due to the presence of the animals so often 



often lined with down, on the lower surface. found therein. 



These afford shelter to insect larvte, to whose J We must recall the very artificial nature of 



agency the production of the pits has been these divisions, especially of such series as the 



ascribed. But this view appears to us untenable, Ocotea, which we only admit to facilitate the 



as we have seen these depressions indicated in study of this most natural order. There is no 



very young leaves of the Camphor-plant while single constant differentiating character, 

 still enveloped in the bud, before they could 



