MINUTE STRUCTURE OF PLANT HYBRIDS. 



245 



each averaging 1^ to 2 /x across. The hair cells are very numerous, uniformly distributed, 

 variable in size, and of conical shape, with rounded apex (Plate VII. fig. 3). In 3 the 

 epidermal cells are as a rule elongated and sharply angular, their chromoplasts are very 

 abundant, of a bright yellow colour, and partly distributed through the peripheral proto- 

 plasmic layer, partly aggregated in many cases round the nucleus. The presence of these 

 gives the yellowish-red colour to those areas of the sepals from which hairs are absent. 

 The hairs are globular or broadly obovate in shape, and though they vary a little in size, 

 the degree of variability is not nearly so pronounced as in 1 (Plate VII. fig. 5). In 2 the 

 number, size, and tint of the chromoplasts, as also the shape and size of the cells which 

 contain them, is a very fair mean between those of the parents, while the pigment hairs 

 are of extreme interest as showing that the variability in size inherited from one parent 

 is reproduced, though on a less exaggerated scale, in the hybrid progeny (Plate VII, fig. 4). 

 Occasionally in the hybrid, as in M. Veitchiana, compound hairs, formed of a short 

 columnar stalk and enlarged head, are observed. 



Stamens. — The cells of the anther sacs in 1 are small and rounded ; of 3 they are 

 elongated and angular ; and of 2 rounded-angular. The pollen grains of 2 are quite 

 equal in quality to either parent, so far as microscopic examination can decide. If, 

 therefore, the ovules are correspondingly good, this hybrid should be capable of per- 

 petuation. 



The suggestions already made as to the evolutionary origin of Lapageria and 

 Philesia might equally be urged here. Both parents are inhabitants of the mountains 

 of South America, and their close histological affinity leads one to look for some common 

 ancestral form. We are now acquainted with several species that are nearly related, and 

 which frequent the same regions, and it is possible that a comparative study of their 

 tissues might aid us in determining the lines of development on which these have 

 advanced, and the possible relation of the artificially produced hybrid to some type which 

 once existed or still exists. 



1 Cypripedium 

 Spicerianum. 



2. Cypripedium 



Leeanum. 



3, Cypripedium 



insigne. 



(i) Cypripedium Leeanum, x . 



This hybrid orchid was raised in Messrs Veitch's Chelsea Garden Nurseries by cross- 

 ing of C Spicerianum with pollen of C. insigne. It is a very evenly-balanced cross on 

 first appearance, but, as we shall show, some parts of it exhibit a pronounced one-sided- 

 ness of development. 



Root. — I have carefully examined roots of the three, but find them fundamentally 

 alike, so far as my supply of material has enabled me to make an exact age 

 comparison. 



Leaves. — Comparative surface views of the upper and lower epidermis show con- 

 siderable differences in the number of epidermal cells and of stomata over a given area. 

 The following results illustrate this. In C. Spicerianum the lower half of the upper 

 epidermis showed, under Zeiss' D with 4 eyepiece, five to six cells, over the upper half 



