78 The Plant World. 



sions are the following. The trichomes vary in size within cer- 

 tain and limited bounds, and in a manner directly associated 

 with the position occupied on the leaf, and not with the differences 

 in vigor of the plants, at least in certain Juglar.s hybrids, al- 

 though, so far as observed, the trichomes on young organs may 

 be larger than on organs which are old. The distribution and 

 reversion of trichomes indicate that hairs should not be grouped 

 together as a "unit" character, but that each type of hair 

 should be separately studied and considered a distinct "char- 

 acter." 



The leading interest in the study lies in tracing the ontogeny 

 of the glandular trichomes in the Juglans hybrids and pure 

 species. The multicellular trichomes were found to have a form 

 of developm^ent which is rigorously adhered to, at least in the 

 smaller hairs, and similar hairs, not identical, were observed to 

 have analogous sequence in cell-division by which they are or- 

 ganized. Where the development of different hairs takes a 

 parallel course, kinship, based on the probability of common 

 ancestors, however remote, is assumed. A deviation from the 

 usual sequence of cell-division was observed in one type of hair 

 by which an aberrant trichome resulted. This sudden de- 

 parture is held to be an actual mutation, so that the trichomes in 

 Juglans appear to take their origin in at least two ways, namely, 

 by slight alterations in the number of cells, as more or less than 

 the customary number, although there might be also a quanti- 

 tative change, or by a new sequence of cell-division,, which would 

 result in a qualitative change. Thus the origin of structures in 

 plants is found by histological studies to agree with the method 

 of their origin as postulated by the experimenters in this field., 



