THE INFLORESCENCE, AND THE FLOWER 265 



however prevalent, especially in highly organised flowers. Since the 

 spiral is characteristic of many primitive flowers and these graduate 

 into cyclic types, the facts suggest that the cyclic state may often 

 have been derived in Descent from condensation of a spiral scheme. 

 Whether arranged in whorls or in spirals, the parts of successive series 

 alternate as a rule. This allows of their being closely packed in the 

 bud. They are formed in acropetal succession. Occasionally it is 

 otherwise, but the most prominent exceptions occur where an organ 

 has been reduced, as in the calyx of the Compositae. 



(ii) Meristic differences. The number of parts in each successive 

 category may differ in different flowers, and these are called meristic 

 differences. In spiral types the numbers are relatively large and in- 

 definite (Figs. 188, 189) ; in cyclic types they are smaller, and usually 

 definite (Fig. 187). Where the whorls are well defined the actual 

 numbers of parts in each may be compared, and are found to vary 

 in different flowers. Some fundamental number ', commonly three, four, 

 or five, then rules in the construction of each flower. Such flowers 

 may be described as trimerous, tetramerous, or pentamerous respec- 

 tively. But the fundamental number rarely holds through all the 

 whorls, though the Flax is an example of this (S 5 , P 5 , A 5 , G 5 ). Usually 

 the androecium shows larger, and the gynoecium smaller numbers 

 than the sepals or petals. 



Not only are meristic differences common between different families, 

 genera, or species, but even between different flowers of the same inflorescence. 

 As regards families, the Crassulaceae show meristic variation in high degree, 

 the fundamental figure rising in the House-leek to as many as twenty, whereas 

 in Sedum it is commonly five. Within the family of the Liliaceae Maianthemum 

 has 2-merous flowers, most Liliaceae have 3-merous, but Paris has 4-merous, 

 or even 5-merous flowers. Within the Primulaceae Glaux sometimes has 

 4-merous, Primula 5-merous, and Lysimachia 6-merous flowers, and others 

 have still higher numbers. Within the genus, Gcntiana campestris has 

 4-merous, and G. amarella 5-merous flowers, while species of Saxifraga may 

 show flowers 5-, 6-, or 7-merous. In the same inflorescence Adoxa and Pitta 

 both show meristic variation. In Rxita the terminal flower is 5-merous, and 

 the lateral flowers 4-merous. In Adoxa, as a rule, the terminal flower is 

 4-merous, and the lateral flowers 5-merous. Such facts are a warning against 

 any undue faith in numbers of parts as themselves indicative of affinity. 



(iii) Fusion of parts. In some simple flowers like the Buttercup 

 all the floral parts are separate, or free from one another. This state 

 is probably primitive, and corresponds to the condition seen in most 

 vegetative buds. But in many flowers certain parts are found to 

 be fused together in the mature state. There is a real continuity of 

 tissue between them. A familiar instance is the Primrose, where the 



