METAMORPHOSIS. 



169 



inrolling of lamellae of a leaf fused to another, or 

 separating off, by its upper surface (as in fig. 4). We 

 thus see how the same anther- structure might be pro- 

 duced by two quite different types of leaf-division (or 

 fusion), such as are shown in figs. 4 and 5. 



Hence, in a case like that of fig. 5, the surfaces of 

 the two fused leaves (PL XLIV, fig. 5) are of 

 necessity in a plane at right angles to the morpho- 

 logically upper surface of the anther. An anther 

 formed in this way would have its marginal loculi 

 situated apparently on the dorsal side ; and such 

 anthers, known as u extrorse," are quite frequent in 

 Nature ; in extreme cases the median loculi are also 

 frequently situated more on the dorsal than the ventral 

 side ; such conditions are merely due to great con- 

 vexity of growth on the upper side of the connective. 

 If a similar convexity of growth were to occur on the 

 dorsal side of an extrorse anther which has arisen 

 from the fusion (or separating off) of two laminae by 

 their dorsal surfaces, then the ordinary introrse 

 anther-structure would result. 



In one of the leafy anthers of Jatropha, viz., that 

 shown in PL XLIV, fig. 4, we see a structure like that 

 in fig. 3, which is brought about by the very intimate 

 fusion of two leaves by their upper (ventral) surfaces 

 forming a single structure. This has dichotomized, 

 and the resultant structure, had we no other clue to its 

 origin, might well be interpreted as the result of fusion 

 of two leaves by their dorsal surfaces, as in fig. 5, and 

 if these became changed into fertile anthers the 

 structure as shown in fig. 128 (p. 170) would result. 



From all of which we see that the two kinds of 

 facial leaf-fusion or -fission above described may have 

 given rise respectively to introrse and to extrorse 

 anther-structure, and these two types of anther, how- 

 ever different they may appear from a study of their 

 ontogenetic development, are yet essentially the same. 



The enations which sometimes occur, as in the 

 Polygonum, along the lateral veins as well as along the 



