102 



MOEPHOLOGY 



tions or perform similar 'labor, without respect to origins, 

 are analogous. Thus a leaf -tendril is analogous to a 

 branch -tendril, but the two are not homo- 

 logous. 



212. There are five tests by means of which 

 we may hope to determine what a given part 

 is : ( 1 ) by the appearance or 

 looks of the part (the least reli- 

 able test) ; (2) by the position 

 of the part with relation to other 

 parts its place on the plant ; 

 (3) by comparison with similar 



147. Leaf and clad- , / 



ophyllaofaspar- P arts On tner Points (compara- 148. Leaves of 



agus. tive morphology) ; (4) by study of 



intermediate or connecting parts; (5) by study of the 

 development of the part in the bud or as it originates, 

 by means of the microscope (embryology). The last 

 test can be applied only by the trained investigator, but 

 it often gives the most conclusive evidence. Even with 



149. Fern-like leaf-branches of a 

 greenhouse asparagus. 



the application of all these tests, it is sometimes im- 

 possible to arrive at a definite conclusion as to the 

 origin or morphology of a part. For example, it is not 

 yet agreed whether most cactus spines represent leaves or 



