92 SCIENCE PROGRESS. 



morphosis, which had hitherto been confined to the leaves, 

 to the other members. For instance, we now know that a 

 tuber, such as that of a potato, or of a Jerusalem artichoke 

 {Helianthus tuberosus), is not a root, though it is sub- 

 terranean, but is a metamorphosed branch ; for if the young 

 shoots at the lower part of the potato-stem be not covered 

 up with earth, as is done in the ordinary course of cultiva- 

 tion, they develop into green leafy branches ; w T hereas, 

 when buried in the soil, they develop into potatoes. 

 Again, we now know that such things as turnips and 

 carrots are really metamorphosed roots. 



But with this extension of the scope of metamorphosis, 

 important knowledge has been gained as to its limits. It 

 may be now laid down as a general law that metamorphosis 

 can only take place within each category of members, and 

 not from the one to the other ; x that, for instance, a leaf 

 cannot be metamorphosed into a stem, or a stem into a 

 leaf, or either of them into a root. At the same time it 

 frequently happens that one member may discharge the 

 functions of another, thus becoming analogous to it ; but in 

 doing so it does not change its morphological nature and 

 become homologous with the member whose work it has 

 taken over. For instance, I would draw your attention to 

 the Cactus {Opuntia) on the table before you. The flattened 

 lobes of which its shoot consist somewhat resemble leaves 

 in form, and actually perform the functions of leaves, yet 

 morphologically they are not leaves, but leaf-like segments 

 of the stem, the plant being leafless. The same thing is 

 even more strikingly shown by this branch of the Butcher's 

 Broom {Ruse us) where the stem, which is of normal form, 

 bears what look like leaves, but are really leaf-like branches 

 {phylloclades). Again, one part of a member may assume 

 the form and discharge the functions of another. Here is 

 a branch of an Australian Acacia in which the stems bear 

 organs resembling leaf-blades ; they are, however, not leaf- 



1 It must, however, be admitted that exceptions do occur. Thus 

 the conversion of the root into a shoot has been observed in Neottia Nidus- 

 avis (see Hofmeister, Allgemeine Morphologie, 1868, p. 428), and in 

 Anthurium longifolium (see Goebel, m-Bot. Zei/g., 1878). 



