MINUTE STRUCTURE OF PLANT HYBRIDS. 261 



as found growing on the " Adami " organism, for I have found the yellow and purple 

 parts of the composite tree to agree exactly with those of the parents, except that the 

 " purpureus " part occasionally shows slight deviations which will afterwards be 

 referred to. 



Stem. — When young twigs of the first year from C. Laburnum and C. purpureus, 

 or shoots of these as found on the composite tree, are examined with the naked eye, they 

 show a smooth green surface, and such is also the case even in four- or five-year-old branches. 

 But while the twigs of "Adami" may remain smooth and shining during the first year, 

 they almost invariably begin to develop a rough, ruptured, and freckled surface in 

 the second year. A simple explanation is got when sections are compared. In C. 

 Laburnum a cork is formed beneath the epidermis towards the close of the first year, 

 and ere long the epidermis peels off as a delicate film. The cork layer is composed of 

 clear, thick-walled refractive cells of a yellowish colour, three to five layers of these being 

 formed each year from the phellogen (Plate VIII. fig. 3). No rupture or peeling off of 

 these may take place for from thirty to fifty years. In C. purpureus a phellogen is not 

 developed till the third or fourth year, and even after that the amount of cork produced 

 is very small, so that the stem is quite encircled by the persistent epidermis, since it 

 attains to no great size (Plate VIII. fig. 1). The strong shoots of C. Adami inherit to 

 a certain degree the vigour of cork formation from C. Laburnum but in a reduced, 

 localised, and retarded manner, while the epidermis has a thick cuticle and shows great 

 persistence as in C. 'purpureus. We find, therefore, that broad but isolated lines 

 or patches of cork develop which eventually rupture the epidermis outside them, and 

 give the rough aspect to the shoots, since they have neither the even uniformity of 

 formation nor of persistence that the cork of C. Laburnum has. Plate VIII. figs. 2 and 

 2 a illustrate stages in the process from a second year's twig, where the irregularity which 

 usually characterises the cork development of C. Adami has been well delineated by 

 my former student, Mr Percy Nicol, to whom I am indebted for several careful 

 drawings. 



In C. purpureus the cortex consists of seven to ten thin-walled cell layers, while 

 in the other two it consists of thirteen to fifteen layers, the cell walls of which are colloid 

 and refractive. But one feature of the cortex calls for special mention. In C. purpureus 

 five longitudinal strands of indurated fibroid elements are imbedded in it at rather 

 irregular intervals (fig. ls.s.), These are entirely unrepresented in the other parent 

 and in "Adami." 



The phloem stereoid masses are of large size, and pretty uniformly disposed in C. 

 Laburnum and C. Adami, but in C. purpureus they vary greatly in size, though they 

 are in all cases smaller than in the first two. The amount and size of the elements of 

 the phloem proper in the first two agree very closely, the zone being relatively broad, and 

 the sieve tubes of large size, as compared with the narrow zone and small sieve tubes of 

 the last. 



As regards xylem, when the average of a series of preparations from different 



