416 DE. M. T. MASTERS ON THE TLOEAL 



That thej did not, liowever, in this case, belong to the inner 

 row was manifest, because the inner row was present m its 

 normal state and position. Moreover, their insertion was dis- 

 tinctly external to that of the fertile stamens, while the back of the 

 column presented a well-marked central ridge concealing a single 

 central vascular bundle, which, near the apex, divided into three 

 divisions — one central, minute, and two lateral, passing into the 

 petal-like shields. The minute central portion evidently repre- 

 Bents the central branch of the filament which in ordinary Orchids 

 bears the anther, wdiile the two lateral lobes are side-outgrowths 

 from the same filament, thus resembling the trifid stamen of 

 Allium^ in which the central division bears the anther. Again, 

 the indications of fission in the median sepal correspond to the 

 similar division of the staminode superposed to it. Indications of 

 this bipartition, or rather tripartition, of the posterior or median 

 staminode are common in many species of Cypripedium, which 

 present a central knob or boss and two lateral lobes. 



On making sections through the peduncle, above the insertion 

 of the bract, a ring of six vascular bundles was seen. At the 

 tase of the ovary the posterior bundle bifurcated; and this twin 

 bundle was traced throughout the whole length of the prominent 

 ridge forming the dorsal suture of Gi, and into the median sepal 

 Si, where the bundles formed the double midrib. A similar 

 section across the column at its base also showed six bundles, 

 five single and one bifurcated \ this bifurcated bundle passed 

 into Ai, its two divisions passing into the shield-like divisions 

 of the staminode. Thus the partition began at the 

 base of the flower in the posterior median line, and Si 

 continued in it all the way up; so that whereas, Ai 

 under ordinary circumstances, the arrangement in Gi 

 the median line would be — L 



in the particular flower now under consideration the arrangement 

 was — S S 



the gynaecium being only partially implicated, and A A 



the lip not at all. 



G^ G 

 L 



Displacement. 



III. C1ia72ge in Fosition of P^r/^.— This has already been 

 incidentally alluded to in connection with the reduced number 

 of parts ; but the importance of the subject necessitates some 

 further mention. In the dimerous forms of Cypripediura the 

 upper or median sepal is usually absent, the lateral ones are 



