1 146 A TEXTBOOK OF THEORETICAL BOTANY 



When only one type of member is present in the perianth the flower is 

 called monochlamydeous, irrespective of whether the members are 

 sepals or petals, though in the vast majority of such flowers the perianth 

 parts are assumed, often without much positive evidence, to be sepals. 

 If petals are present in some altered form, for example as nectaries, which 

 in eflect removes them from the category of perianth members, the flower 

 may still be monochlamydeous and the perianth members are obviously 

 sepals, but where no such morphological indication exists the vascular 

 anatomy may be the only evidence available as to the nature of the existing 

 parts. CaUha is a case in point, the perianth parts being coloured and 

 petaloid in texture but having the characteristic 3-bundle trace of sepals 

 instead of the single trace of true petals. 



Where all the perianth parts are alike the flower is called homoio- 

 chlamydeous, a term often treated as synonymous with monochlamydeous, 

 though it is not so in strict usage, for there are many flowers, especially 

 among the Monocotyledons, where both sets of members are presen, 

 though both are alike in appearance. In the Liliaceae both sets are petaloid, 

 but both may conversely be sepaloid, as in Ritmex. 



Flowers with both sepals and petals present are called dichlamydeous, 

 and if the two categories of parts are dissimilar they are, further, hetero- 

 chlamydeous. 



We have previously referred (p. 1137) to the phenomenon of concres- 

 cence between floral parts as aflPecting the question of the origin of epigyny 

 and perigyny. While it remains doubtful, in many cases, whether a floral 

 cup surrounding the gynoecium has originated by concrescence or by 

 receptacular upgrowth, there are a few hypogynous flowers, such as 

 Hyacinthiis and Asparagus, in which a definite concrescence of sepals and 

 petals has evidently taken place, usually also adding, as in the above examples, 

 concrescence with stamens as well, so that all the whorls, except that of the 

 carpels, may be adherent. 



Although the placing of perianth parts around the floral receptacle 

 follows the general principle of equidistance, to which we have already 

 referred, there is considerable variety in their precise relations to each other. 

 The spiral order has been very widely and perhaps rather uncritically 

 accepted as the most primitive, that is to say, as the order which probably 

 prevailed in the immediate ancestors of the Angiosperms. This view is 

 admittedly bound up with the assumption that the Angiosperms have 

 originated from gymnospermous ancestors with cone-like flowers of many 

 sporophylls, spirally arranged, a view which, though probable, is by no 

 means proved. 



A sounder and more factual approach to the question lies in the com- 

 parative examination of the flowers themselves, and Salisbury, from exten- 

 sive observations of meristic variation in the Ranunculaceae and Alismaceae, 

 has suggested that the primitive condition is one of trimery, that is with 

 parts in whorls, or closely placed groups, of three. This condition, he 

 considers, is related to the mode of cell-division in the meristematic apex, 



