732 THE AMERICAN NATURALIST. [VoL. XXXVI. 
far as sequence is concerned, there appeared to be no definite 
order, series of transition stages running in both directions. 
There was also found a fairly large number of stages which 
lead one to doubt greatly whether in all cases it is the petiole 
only which is transformed to the phyllode, and whether, if one 
pair of pinnze persists at a certain stage, that pair is always the 
proximal one. The prints which follow may, to be sure, repre- 
sent mere anomalies, but from their number, at least, they can- 
not but raise in one's mind a certain amount of hesitation to 

Fic. 4. Fic. 5. 
Fics. 4 and 5. — A. heterophylla, variation in transition to phyllode. (l4 natural size-) 
consider the existence of a law as to method in any way estab- 
lished. Here the flattening appears in some cases entirely on 
the distal portions without affecting the petiole, in others both 
petiole and rachis are involved to varied extents. How these 
are to be interpreted under one definite law seems incompre- 
hensible. 
It is hardly to be expected, however, that the process of 
réduction of parts is going to follow the same strictly regular 
line as that of advancement. The two are brought about by 
entirely different causes and are in no way comparable. There- 
fore the fact that such changes as those last described occur 
does not greatly invalidate the testimony of the earlier presented 
