210 
THE .JOURNAL OF ROT ANT 
imbricate. Subimbricate and subconvolute are ambiguous, as strictly 
they mean “somewhat imbricate” and “somewhat convolute” 
respectively. To avoid this confusion, new and distinctive terms for 
the second and third phases are here proposed. As the essential 
difference between them is in the relative positions of the exterior 
members, the term apotactous («7 to and Tan-os —arranged apart) is 
suitable for the second phase, and paratactous (napa and tciktos — 
arranged side by side) for the third. 
i n ni n 
In the quincuncial diagram the members have been numbered 
1 to 5 in the usual sequence. For purposes of comparison, corre¬ 
sponding numbers have been used in the other diagrams. 
The direction of overlap shown in the four diagrams is clockwise 
(I and IV) or prevailingly so (II and III). By drawing the mirror- 
image of each diagram the corresponding counter-clockwise series 
may be obtained. 
Henslow showed that the quincuncial, “ half-imbricate ” (apo¬ 
tactous), “imbricate” (paratactous), and convolute phases formed a 
sel’ies in which each phase could be derived theoretically from the 
preceding by a single reversal of overlap (Trans. Linn. Soc. ser. it. 
Bot. i. 182). Undoubtedly this is so, but in any such series evolution 
might have taken place in either direction. He offered no satisfactory 
evidence in support of his view that the quincuncial phase was 
primitive relatively to the others. This view seems to have been 
based on his previous hypothesis that f- phyllotaxy in Dicotyledons was 
originally derived from an opposite-decussate arrangement (l. c. 41). 
Evidence for or against the evolution of one phase of overlapping 
aestivation from another may be sought in those species in which 
the {estivation is variable. Henslow examined the aestivation of the 
petals in Ranunculus bulbosus, Viburnum Tinus, Primula vulgaris , 
and Ribes coccineum, in each of which the aestivation varies, and 
indicated the relative frequency of the various phases. But his 
percentages are of little value, as he himself admitted ( l . c. 177), as 
thev are based on examination of an insufficient number of flowers— 
125 of V. Tinus , 120 of P. vulgaris , and an unspecified but probably 
smaller number of Ranunculus bulbosus and Ribes coccineum. His 
figures for R. bulbosus were: 38 per cent, quincuncial, 14 per cent. 
“ half-imbricate ” (apotactous), 32 per cent. “ imbricate proper ” 
(paratactous), and 12 per cent, convolute. They give a misleading 
impression of the relative frequency of the various phases. The 
