2 Anatomie. 



tales. It is pointed out that in the phyllode theory the Coniferae are 

 microphyllous by reduction. 



The phyllode theory has hitherto been based entirely on exter- 

 nal morphology, but in the second part of the present paper reason 

 is given — on the ground of a comparison of Dicotyledonous scale- 

 leaves, petioles, and phyllodes with the leaves of Monocotyledons — 

 for the view that the occurrence of inverted vascular bundles, to- 

 wards the adaxial face of a leaf, may be an indication of phyllodic 

 morphology. A list is added of the eases of such structure in 

 Monocotyledons. 



The isobilateral equitant leaf of Iris, etc., is regarded b} T the 

 present writer, not as a exhibiting congenital concrescence of the 

 two halves of the organ, but as a special case of phyllodic 

 structure. It is shown that this type of leaf is widely distributed 

 among the Monocotyledons, that it can be closely compared with 

 certain Acacia phyllodes, and that transitions can be traced between 

 it and other types of Monocotyledonous phyllode. 



Anatomical evidence is brought forward in favour of Hens- 

 low's corollary to De Candolle's theory. It is shown that the 

 'laminae' of certain Pontederiaceae {Eichhornia, Pontederia, Heteran- 

 thera) all agree in the presence of inverted as well as normal 

 bundles. Inverted bundles are also recorded in the 'lamina' of Sa- 

 gütaria. Attention is drawn to Solereder's discovery of inverted 

 bundles in the 'laminae' of certain Hydrocharitaceae. 



The systematic distribution of phyllodic anatomy among the 

 Monocotyledons is dealt with, and it is concluded that the systema- 

 tic distribution harmonizes with the view that the type of anatomy 

 here called 'phyllodic' is an ancient character, revealing the petio- 

 lar origin of the Monocotyledonous leaf. 



Agnes Arber (Cambridge). 



Bailey, J. W. and W. P. Thompson. Additional Notes 

 upon the Angiosperms Tetracentron, Trochodendron , and 

 Drimys, in which V esseis are absent from the Wood. 

 (Ann. Bot. XXXII. p. 503-512. 1 pl. 9 text figs. 1918.) 



The writers in a former paper (Mem. N. Y. Bot. Garden, VI, 

 1916, p. 27 — 32) showed that true vessels are entirely absent from 

 the three genera Tetracentron, Trochodendron and Drimys. The 

 accuracy of these conclusions was questioned by Jeffrey and Cole 

 (Ann. Bot. XXX, 1916, p. 359—368) who claimed to have found 

 vessel-like structures in injured roots of Drimys colorata, and who 

 concluded that the ancestors of these three genera possessed true 

 vessels. They admitted, however. that the elements described and 

 figured by them lack the perforations of normal vessels. The pre- 

 sent writers have reinvestigated these elements and conclude that 

 they are typical tracheides having transitional t}<pes of pitting, such 

 as occur in many Pteridophyta, Gymnospermae and Angiospermae. 

 They lind that they occur in uninjured stems and roots of Tetra- 

 centron, Trochodendron and Drimys, as well as in traumatic tissue. 

 Not only is there no direct struciural evidence which might be con- 

 sidered to indicate that these vessel-less Angiosperms are 'evascula- 

 rized' forms, but there appears to be no physiological and ecologi- 

 cal evidence for supposing that they are 'degenerate' Dicotylcdons 

 whose ancestors possessed true vessels in their secondary wood. 

 The secondarv xvlem of Tetracentron and Trochodendron, which is 



