POSITIVE D^DOUBLEMENT. 



65 



median sepal necessarily tends to disappear in obedi- 

 ence to tlie law of alternation. Where, however, the 

 original arrangement of the various floral members 

 obtains, there is some likelihood of the reappearance 

 of this third sepal, as was observed by Goebel in the 

 development of the flower of a species of Streptochseta 

 (fig. 85), the mature flower (fig. 84) showing no trace 

 of it, and by Schacht in that of the flower of Agro- 

 pyrum giganteum (Triticum rigidum). 



An interesting case is described by Velenovsky in 

 the male flower of Salix aurita, viz., the reappearance 

 of the median anterior perianth-leaf. In the normal 



84 85 



Pia. 84. — Streptochseta. Section of mature flower. (After Goebel.) 



Pio. 85. — Streptochseta. Young flower showing presence of anterior 

 sepal (j)'). p', p'', posterior sepals. Ax, axis of spikelet. (After 

 Goebel.) 



flower of S. pentandra and some others this organ is 

 also present in the form of a gland. But in the allied 

 genus Populus this anterior perianth-leaf is normally 

 present fused with the bracteoles to form the cupule- 

 like structure. In the allied orders Juglandacese and 

 Myricaceae the two medianly-placed perianth-leaves 

 are normally present as very small leaves. In two 

 strongly virescent flowers of 8. aurita the gland 

 assumed a scale-like form. 



Schmitz observed, in the abnormal cypress-spurges 

 (Euphorbia) he examined, the appearance of one or two 

 small leaves immediately below the anthers of the 

 axillary " stamens," and also immediately below the 



VOL. ir. 5 



