34 TUE JOURNAL OF BOTANY 



SO in latifolla) and are longer, more acute, and much more acuminate 

 than in the latter. The epichile is long, narrow, and acuminate, not 

 broad and recurved as in latifolia. 



Differences in the Eepeoductite Organs. 



Plate 553 shows more clearly than a detailed description the 

 differences between the two species. In figs. A 1 (^. viridiflora) 

 and B 1 {E. latifolia) the following points should be noted : — In A 1 

 the ovary is si3indle-shaped, broadest about the middle, tapering 

 gradually towards each end. In B 1 it is pear-shaped, thickest near 

 the apex, and somewhat hump-backed. Both the basal cup and the 

 terminal lobe of the lip differ in shape, the latter being long and 

 narrow in A 1, shorter, broader, and recurved in B 1. The column in 

 Al is divided into two parts, but undivided in B 1. The anther 

 in A 1 is very decidedly stalked, slender, acute, and projects well over 

 the upper edge of the stigma ; in B 1 it is sessile, thicker, more 

 obtuse, and does not project over the stigma at all. In Al the 

 poUinia project over the stigma, in B 1 they are entirely behind it. 

 In A 1 the rostellum has disappeared (the flower having been open 

 some hours) ; in B 1 it is large and persistent. 



In A 2, a back view of the column of viridiflora, the stalk on 

 which the anther is perched, and the tapering of the anther to an 

 acute point, contrast strongly with B 2, a similar view of latifolia. 

 A3, a front view of the column of viridiflora, shows how the 

 poUinia (/?.) emerge over the upper edge of the stigma, forming a 

 wide inverted V or horse-shoe, with a dark spot (r.) in the apex, the 

 withered remains of the rostellum ; the projection on the right is an 

 abnormally developed staminode {st.), within which a yellowish rudi- 

 mentary poUinium (;;.) could be seen. B 3, a similar view of latifolia, 

 shows the large persistent rostellum. CI, C 2, and C 3 show similar 

 details of E. violacea. 



These figures were drawn from living specimens by my wife, who 

 at the time was not familiar with the differences between the species, 

 and simply drew what she saw. They are not therefore open to the 

 objection of unconscious exaggeration by an artist with preconceived 



ideas. 



It will be seen from the above that there are marked differences 

 both in the vegetative and in the rej^roductive organs of the plants. 

 Whilst variations in size and shape occur in the leaves, bracts, and 

 perianths of orchids, often amounting to several, or even to many 

 millimetres in extent, without affecting prejudicially the welfare of 

 the plant, the organs of reproduction themselves ai-e fashioned with 

 the utmost exactitude. Every part must be faultless in form and 

 dimension, and all the parts co-ordinated with absolute precision. 

 Deviations which would be unimportant m other parts of the plant 

 would here be fatal, and throw the delicate mechanism out of gear, 

 causino" the plant to fail to perpetuate its kind. For this reason 

 variations in the structure of the reproductive organs are relatively 

 of much greater importance than any that occur in the vegetative 

 parts of the plant, for while the lattt'r can only affect the individual, 



