172 



From the inner end of Canary Avenue, in front of the palace 

 of the governor-general, a formal avenue of royal palms leads to 

 another gate. From these avenues as a base, other walks and 

 drives extend through the whole garden, always past plants of 

 great botanical interest, but never past any single scene with as 

 much attraction for the ordinary tourist as the famous Canary 



Avenue. 



The botanist can easily spend a whole day along this avenue 

 alone, examining the lianas and epiphytes. Among the aroids, 

 several species of our common greenhouse genus Philodendron 

 are planted, as well as various other genera seldom seen in culti- 

 vation in America. Few of them have developed the long 

 pendent aerial roots, so frequently seen in the Philippine forest, 

 although they are all root-climbers. In many cases roots have 

 reached the ground from somewhere along the stem, so that the 

 stem itself is no longer essential. The stem of one big Philo- 

 dendron, four inches in diameter, was completely dead and severed 

 from the ground, while for two or three feet the wood was eaten 

 away by termites until only the bark remained. Still a few of 

 these roots reaching the ground were sufhcient to keep the plant 

 in a flourishing condition. 



From a physiological standpoint at least, there are two kinds 

 of these aerial roots. One serves for holdfasts only, and such 

 roots are diageotropic, or nearly so. If there is any deviation 

 from the horizontal position around the tree-trunk, it is usually 

 upward, making them slightly negatively geo tropic. The whole 

 root, even to the extreme tip, is closely appressed to the bark, 

 indicating a strong thigmotropism, although their uniform direct- 

 tion, not influenced by irregularities in the bark, indicates that 

 the thigmotropic response is subordinate to their diageo- 

 tropism. 



The second kind of root is always positively geotropic and 

 sends out absorbent roots if it finally reaches the ground. These 

 roots are also strongly thigmotropic, and two of them may some- 

 times parallel each other around all sorts of crooks and turns for 

 three or four feet. In this case the thigmotropism seems to 

 outweigh the geotropism, for a root may make abrupt turns to 



