Doyle — Some Researches in Experimental Morphology . 421 



Thus, Plate XXXII, fig. 4, is a section of a well-developed petiole of 

 Solaimm Balbesii. The single bow-shaped bundle is prominent ; (1) is the 

 upper, (2) the lower, surface. 



Fig. 3 shows very diagrammatically the result of grafting the sprout on it. 

 The old cambium awakens and forms a mass of secondary wood {b). But as 

 well as in tlie region marked («) in Plate XXXII, fig. 4, a cambium appears 

 which gives rise to the heavy mass of wood on the upper side of the petiole. 

 The two cambiums unite at e and e, and in this way a complete cylinder 

 of wood is formed. The wood from the new cambium presents peculiarities 

 which unfortunately have not yet been studied. But at places such as/ and 

 / we find not normal secondary wood but very large tracheides, so large as to 

 appear not like tracheides cut across, but as if the tracheides had tlieir 

 longitudinal increase horizontally instead of vertically. It was as if the 

 tracheides had been developed in layers — here a column of vertical tracheides, 

 against it a column heaped high of tracheides lying horizontally. Of course 

 we might expect tracheides varying from the normal as they are derived 

 from a cambium which has abnormally arisen in large differentiated cortical 

 cells ; but why one part should give rise to normal secondary wood, and 

 another part to this formation, seems difficult to explain. It might be 

 suggested, however, that it is in some way a response to mechanical stimuli, 

 as we shall later discuss. As to its mechanical efficiency, no opinion, of course, 

 can be expressed. 



Fio 3. — Solanum Balbesii. Old grafted petiole. Diagrammatic. 



(a) is a thick cork layer developed where the outer layers burst with the 

 internal expansion. The cortex at (b) is still intact. 



The increase in diameter is marked. Thus the normal petiole measures 



SCIENT. PEOC. R.D.S., VOL, XIV., NO. XXXIII. 3 Y 



