130 VARIOUS OBJECTIONS. Chap. IV. 



reproductive organs are frequently more or less aborted. It is 

 a more curious fact that the achenes or seeds of the circum- 

 ference and of the centre sometimes differ greatly in form, 

 color, and other characters. In Carthamus and some othoi 

 Composita^ the central achenes alone are furnished with a pap- 

 pus ; and in llyoseris the same head yields achenes of three 

 different forms. In certain Umbcllifeni:; the exterior seeds, ac- 

 cording to Tausch, are orthospermous, and the central one 

 coelospermous, and this difference has been considered by De 

 Candolle as of the highest systematic importance in the family. 

 If in such cases as the foregoing all the leaves, flowers, fruits, 

 etc., on the same plant had been subjected to precisely the 

 same external and internal conditions, all no doubt would have 

 presented the same morjohological characters ; and there clearly 

 woiild have been no need to call in the aid of the principle of 

 progressive development. With the minute closed flowers, as 

 well as ^^^tll many degraded parasitic animals, if it be assumed 

 that any such aid is requisite, we sliould have to call in an in- 

 nate tendency to retrogi-essive development. 



]\Iany instances could be given of morphological characters 

 varying greatly in plants of the same species growing close 

 together, or even on the same individual plant ; and some of 

 these characters are considered as systematically important. 

 I will specify only a few cases Avhich have first occurred to me. 

 It is not necessary to give instances of flowers on the same 

 plant being indifferently tctramcrous, pentamcrous, etc. ; but 

 as when the parts are few, numerical variations are in all cases 

 comparatively rare, I may mention that, according to De Can- 

 dolle, the flowers of Papavcr bracteatum offer two sepals with 

 four petals (and this is the common type with jioj^pies), or 

 three se}>als with six petals. The manner in which the petals 

 are folded in the bud is in most groups a constant morpho- 

 logical character ; but Prof. Asa Gray states that with some 

 species of Mimulus, the aestivation is almost as frequently that 

 of the Rhinanthidea3 as c>f the Antirrhinidea?, to which tribe 

 the genus belongs. Aug. St.-Hilaire gives the following cases ; 

 the genus Zanthoxylon belongs to a di\-ision of the Kutacea3 

 with a single ovary, but in some of the species flowers may bo 

 found on the same plant, and even in the same panicle, with 

 cither one or two ovaries. In Helianthennnn the capsule has 

 been described as unilocular or 3-locular ; and in H. nuitabih^ 

 " Une lame, plus ou moins large, sY'tend entre le pericarpe et 

 Ic placenta." In the flowers of Saponaria ollicinalis, Dr. Mas- 



