PETALOIDE/E. 371 



rarely a fleshy indehiscent pod ; seeds very numerous and ex- 

 tremely small, consisting of a cellular nucleus without distinct 

 radicle or plumule, enclosed in a loose membranous or rarely 

 crustaceous testa. 

 The Order has been divided by Lindley, its greatest exponent, into 



several tribes, according to the number and position of the anthers, the 



number and nature of the pollen-masses, &c. 



1. Anther solitary. 



A. Pollen-masses waxy. 



a. No caudicle or separate stigmatic gland . . Tribe MALAXEJE. 



b. A distinct caudicle, but no separate stigmatic 



gland EPIDENDREJE. 



c. A distinct caudicle and stigmatic gland .... YANDEX. 



B. Pollen powdery, granular, or sectile. 



a. Anther terminal, erect OPHRYE^E. 



b. Anther terminal, opercular ABETHUSE^E. 



c. Anther dorsal NEOTTEJE. 



2. Anthers two CYPRIPEDIE.E. 



Affinities, &c. In the greater part of the genera the Monocotyledo- 



nous type is departed from in several particulars, as : in a more or 

 less considerable irregularity of the perianth, especially in the condition of 

 the labellum ; in the circumstance that the filaments are confounded with 

 the style into a central organ, prolonged from the inferior ovary, called the 

 column, and that generally 2 out of 6 (at least 3) anthers, are abortive, 

 while the pollen is frequently less developed than usual, the process of 

 subdivision into distinct cells or granules being arrested, so that it remains 

 in compound masses of various degrees of magnitude and of more or 

 less firm and even waxy consistence. In some cases, however, as in 

 Thelymitra, the perianth is almost regular, so as to resemble that of 

 some of the genera of Iridacese ; and in Cypripedi&m we find 2 anthers 

 developed and the rest abortive. Among the other remarkable peculiarities 

 of the structure are processes of various kinds occurring upon the column 

 and labellum, which there is reason to regard as indications of abortions 

 of staminal organs. These have given rise to the opinion that the ele- 

 ments of 2 circles of stamens exist in this Order, of which 5 are usually 

 suppressed, the perfect one belonging to an external circle of 3, while in 

 Cypripedium the 2 which are developed are members of the inner circle 

 of 3, one of which, with the entire outer circle, is abortive. The position 

 of the organs may be thus shown : In the typical flower the arrange- 

 ment would be S . . P . A .-. + . U . ' . In most Orchids the ar- 

 rangement is | S.'.P'.'A * 4- G . . , the dots representing those 

 organs that are present and their relative position, the circles those organs 

 that are suppressed. In Cypripedium the formula is S . . P . A 

 -f- ' (Jf . . These views are supported by numerous exceptional in- 

 stances, in which some or all of the ordinarily suppressed stamens are 

 present, and by the anatomical construction, which reveals the existence 

 of as many bundles of vascular tissue in the column and ovary as there 

 are stamens and carpels. Thelasis has normally three stamens. The 

 suppression of 2 out of 3 stamens connects this Order with Marantaceae 

 and Zingiberaceae, where the same phenomenon exists in a different modi- 



2 B2 



