128 ALTERATION OF POSITION. 



similarly detached from the ovary simultaneously with 

 the extension of the axis. Here the petals were in- 

 creased in number and variously modified, the stamens 

 also ; while in the centre and at the top of the flower, 

 conjoined at the base with some imperfect stamens, 

 was a carpel open along its ovuliferous margins. Such 

 instances as these seem to be the first stages of a 

 change Avhich, carried out more perfectly, would result 

 in the formation of a new bud on the extremity of the 

 prolonged axis. 



In Orchidacea\ among which family I have now met 

 with several instances of prolification, the ovary seems 

 usually to be absent. Fig. 63 shows a prolified flower 



Fig. 63. Median prolification in Orchis j^yromidalut, the ontev seg- 

 ments of the perianth regular and reflexed. 



of Orchis pyramidalis in which the perianth was nearly 

 regular, the central portions of the flower absent, and 

 their place supplied by a new miniature raceme. This 

 specimen was forwarded to me by Dr. Moore, of 

 GlavSnevin. 



As might be expected, it very rarely happens that 

 median prolification occurs without some other deviation 

 in one or more parts of the flower being simultaneously 

 manifested. Some of these changes have l)een already 

 mentioned, but others are commonly met with, as, for 

 instance, the multi})lication or doubling, as it is termed, 

 of the petals ; others, though less fretjuent, are of more 



