168 RELATION OF PHYLLOTAXIS TO MECHANICAL LAWS. 



DipsacMs fullonum, having been very fully investigated by Bravais, 

 may be taken as a type of the bijugate condition. Seedlings flower 

 in the second summer, and the plants usually die after fruiting ; 

 the seedUngs of the first year form a tuft of leaves with a very 

 definite spiral arrangement. In other species, D. sylvestris, 

 D. laciniatus, a weU-marked radical-rosette is produced, in the latter 

 the aggregate of leaves being flattened out on the soil to form a 

 rosette two feet in diameter, in which apparently no two leaves are 

 superposed, and to aU appearances the spiral construction is that of 

 the normal series (fig. 61&). If the plant be cut across (fig. 61a), the 

 contact parastichies are seen to be well marked in the bud and 

 constantly (2 + 4); the first year's plant being thus bijugate as a 

 seedling without any apparent reason. 



In the second year a tall leafy shoot is sent up which bears leaves 

 of the specialised "bucket" type, most marked in J), laciniatus; 

 this shoot is at first sight symmetrical with " decussate " phyllotaxis, 

 and beyond the vegetative leaves the apex produces a complex 

 terminal capitulum which in well-nourished plants is practically 

 constantly bijugate of the type (26-1-42). 



Thus Bravais found this type in 272 out of 350 capitula, or over 

 77 per cent. In the progressively smaller lateral heads, other 

 systems appear, often trijugate, but sometimes of the normal series, 

 and equally often anomalous systems or quite undeterminable types 

 occur. Bravais tabulates 4 per cent, normal series, 4'5 trijugate, 7 

 per cent, undeterminable, and 6 per cent, anomalous systems. The 

 facts, then, show that Di^psouius presents an example of a plant with 

 a specialised leafy axis, springing from an asymmetrical system, and 

 exhibiting, when the vegetative period is over, another asymmetrical 

 system which, like the first, is normally bijugate. 



The construction of a Dipsacvs plant, then, is very remarkable if 

 these facts are true, — that it commences with a (2-|-4) rosette, 

 becomes symmetrical (decussate) in the leafy shoot, and then 

 produces a bijugate inflorescence (26+42); since this would imply 

 that a double transition from asymmetry to symmetry takes place in 

 the life of the plant in passing firstly from leafy rosette to tall leafy 

 shoot, and secondly, from inflorescence to flower. The assump- 

 tion of symmetry in the floral members is so general that it 



