folia. 



218 DR J. M. MACFARLANE ON THE 



/. Lapagena >phe m esophyll substance agrees with that of root, stem, and vegetative leaf in that 



8. Phiiagena its cells are smallest in 1, largest in 3, and intermediate in 2 ; but, further, Philesia has 



Veitchn. ° 



* P iv!if,? ia buxi " brown pigment cells of varying size, which are well shown in fig. 10. I have failed to 

 find traces of these in the hybrid, though we shall see that they appear in the petals, 

 beiuij inherited from those of Philesia. 



Petals. — In 1 the honey-gland is developed at the bottom of a rather deep pouch, the 

 opening of which is seen between the bases of the stamens, and these cover the gland cavity 

 over two-thirds of its area. In longitudinal section its upper part protrudes suddenly from 

 the petal surface, and forms a thick pad which gradually tapers out below. At its upper 

 thickest portion it consists of nineteen to twenty-one cell layers deep. The vascular bundles 

 distributed to the gland are arranged in a rather loose and open manner, one set of bundles 

 — the gland bundles proper — lying immediately beneath the gland tissue, the remainder, 

 from which the former are given off, lying external to, and in most cases alternate with 

 them. In 3 the gland is a nearly circular pad of tissue lying exposed at the base of the 

 petal, and though on longitudinal section it swells out rather abruptly above, it retains a 

 very uniform depth throughout till near its base. The greater mass of the gland is 

 composed of ten to eleven cell layers, and beneath the whole the vascular bundles are 

 densely arranged side by side, or obliquely beside each other. In 2 the gland is in every 

 respect intermediate, for I have failed to find any feature in which it specially sways to 

 either parent. 



The external epidermal cells in the three greatly resemble each other, but when one 

 turns to the inner surface those over the upper part and down the middle towards the 

 base of the petal in 1 are mostly broader than long, though at the edges they are 

 equilateral, or longer than broad. In 3 the cells at the top are mostly twice as long as 

 broad, while below they become even more elongated proportionately. The hybrid cells 

 are intermediate. One does not find cells in Lapageria with brown granular contents 

 such as occur on the sepals and petals of Philesia : but though absent on the sepals so far 

 as I can trace, they are present in the petals of the hybrid, though the contents are paler. 



Stamens. — I have not attempted to compare minutely either stamens or carpels, but 

 the pollen grains have been examined with care. In both parents the pollen cells are 

 spherical, echinated, and well formed in good flowers, those of Philesia being slightly larger 

 than those of Lapageria. The cells of the hybrid, however, are to all appearances bad 

 to the extent of at least 95 to 98 per cent. They are very irregular, small, shrivelled, and 

 starved-looking shells ; a few approach in size and form to those of the parents, but only 

 two or three in a hundred are at all well formed. I should consider, therefore, that 

 attempts to recross this hybrid would be attended with great difficulty. No com- 

 parison is made of the more vital parts in Philageria or its parents, since the very 

 pronounced character of their cell walls militate against convenient examination of the cell- 

 contents ; but the careful manner in which moleeule has been added to molecule in these 

 walls, proves that the building substance or protoplasm which has accomplished this work 

 in the hybrid must be an intimately blended product of that of the parents — i.e., that 



