168 THE STRUCTURE OF FLOWERS. 



dissoudre la cellulose, ainsi que le prouvent les soudures avec 

 fusion que j'ai observees plusieurs fois entre eux dans les 

 cultures, ou le phenomene est plus facile a voir. D'ailleurs, 

 la penetration directe des tubes polliniques dans les papilles 

 du stigmate de plusieurs fleurs, apres dissolution de la 

 membrane cellulaire, est un fait du meme ordi'e." 



I quote this passage in full, that the reader may see how 

 it comjjletely corroborates my belief that the metamorphosis 

 of the epidermis and subjacent layers to form the conducting 

 tissue is entirely owing to the action of the tubes themselves, 

 as well as the conversion of starch into saccharine, and there- 

 fore easily absorbable matters. 



M. P. Maury has noticed very analogous facts in Ve7-bas- 

 cum, in that " at the period of pollination the ovules are still 

 in a rudimentary condition, and altogether unfit for fertilisa- 

 tion. The nucellus is entirely occupied by the embryo-sac, 

 in the protoplasmic contents of which there is as yet no 

 differentiation of oosphere, synergidee, or antipodals. It is 

 only after the pollen-tube reaches the micropylar canal that 

 these begin to be formed." * 



This observation corroborates what I have said above, 

 that not only is the pistil delayed in development in insect- 

 crossing flowers, but that arrest of growth may affect all 

 parts, and particularly the ovules ; and I strongly suspect if 

 more instances, of the Gamopetalce especially, were examined 

 it would be found to be the rule and not an exception. M. 

 Maury's investigations also agree with M. Guignard's, in 

 that the action of the pollen-tube is a stimulating one, and 

 brings about developments which would not, and, indeed, 

 cannot, otherwise take place. 



In Vanda tricolor pallens, experimented upon by M. 



* Bull. Soc. Bot. Fr., viii. (1880), p. 529, quoted from notice in 

 Joum. Roy. Micr. Soc, 1887, p. 433. 



