513 



nmbcllule is composed of two, three, or four flowers. 7. But the 

 most remarkable fact is, that in suiidiy of these peripheral clusters, 

 resulting from the metamorphosis of sim])le umbels into compound 

 umbels, the like metamorphosis is carried a stage Jiigher. Some of 

 the component rays, are themselves the bearers of compound umbels 



€ 



instead of simple umbels. In Fig. 70, a portion of the dried speci- 

 men is represented. Two of the central umbellules are marked a 

 and h ; those marked c and d are mixed clusters ; at e and / are 

 compound umbels replacing simple ones ; and g shows one of the 

 rays on which the over-development goes still further. 



Does not this evidence, enforced as it is by much more of like 

 kind, go far to prove that foliar organs may be developed into axial 

 organs? Even were not the transitional forms traceable, there would 

 still, I think, be no other legitimate interpretation of the facts last 

 detailed. The only way of eluding the conclusion here drawn, is by 

 assuming that where a cluster of flowers replaces a single flower, it 

 is because the axillary buds which hypothetically belong to the 

 .several foliar organs of the flower, become developed into axes ; and 

 assuming this, is basing an hypothesis on another hypothesis that is 

 directly at variance with facts. The fohar organs of flowers do not 

 bear buds in their axils ; and it would never have been supposed 

 that such buds are typically present, had it not been for that 

 mistaken conception of " type " which has led to many other errors 

 in Biology. Goethe writes: "Now as we cannot realize the idea 

 of a leaf apart from the node out of which it springs, or of a node 

 without a bud, we may venture to infer," &c. See here an example 

 of a method of philosophizing not unconunon among the Germans 



