1922] BUCHHOLZ—VASCULAR PLANTS 281 
form of biological economy, another feature in which developmental 
selection excels. 
A further objection which has been made to natural selection 
has to do with the difficulties that are involved in explaining how 
the first steps in any given variation may be of selective value. 
How can natural selection influence a structure whose advantage 
is to be reached only at some future time, after the results of the 
selection are achieved? This is asking natural selection to pass 
on a prophecy. Developmental selection, however, is a form of 
selection which can act on some very small quantitative characters. 
For example, by playing on such features as minute differences in 
suspensor length and rate of growth in gymnosperms, or a rapid 
pollen tube penetration in seed plants, developmental selection 
brings about a positive selection for minute differences in these 
particular characters. It may be asked how this selection for 
embryonic or pollen tube vigor could in any way affect a selection 
for other characters. The answer is found in linkage of characters. 
A sporophyte character of the mature plant must be linked with 
the factors producing either vigorous suspensors and embryos, 
or vigorous pollen tubes in the gametophyte stage. 
CorRRENS found just such a case in Melandrium, which was 
referred to previously, in which the female-determining pollen tube 
has the quality of growing slightly faster than the male-determining 
pollen tube. Judging from his account, this linkage is probably 
not an exceedingly close one, as he states that there are some 
male-determining pollen tubes that grow faster than the slowest 
female-determining pollen tubes. Such a condition may be due 
to the well known crossing-over phenomenon associated with 
linkage, between the sex factor and the gene or genes producing 
rapid pollen tube growth. If factors affecting the rate of pollen 
tube growth should become linked with quantitative factors, 
which are usually multiple factors affecting the size of an organ 
or part, it is easy to understand how developmental selection can 
play on them indirectly through their linkages. An accumulation 
of the effect of this selection would account for the building of a 
new structure. Of course natural selection can also play on such 
a character or structure when it or a factor linked with it has become 
