A BLUE BOOK OF RAKE GLADS 
11 
that I extended my operations to the full limit my time per¬ 
mitted and, for the past few years, my annual crop of new 
seedlings has been in the neighborhood of 100,000 
From the very beginning of my work with glads I have 
made it a point to keep systematic, accurate records of cross¬ 
es, and results of each cross, and my record of crosses isi my 
constant companion in all of my breeding work; in fact, 
except when using two named varieties, I never make a crosis 
without first having looked up the full genetic history of each 
parent In this connection I might mention that, while I have 
not excluded the use of standard varieties in my work, I have 
limited the employment of them very greatly for the reason 
that I find my seedlings provide a vastly more promising 
array of material, with the added advantage that I know their 
full genetic history for several generations back. During the 
past season I had provided myself with ample stocks of 
several of the much wanted New Zealand novelties, expecting 
to use them extensively: while I had enough material to keep 
me busy for several weeks, had I chosen to use it, I actually 
did not use them for more than a dozen crosses because they 
did not appeal to me. In other words, in choosing my material 
I am not awed by either an imposing name or a staggering 
price: what I am looking for is some trait that promises to 
be useful for the development of better varieties and my ex¬ 
perience has been that such traits are much more likely to be 
found among my tens of thousands of seedlings than among 
a" few dozens of (usually) much over-advertised novelties. 
While it is possible to get outstanding seedlings from a 
single direct cross, I think that such result is the exception, 
rather than the rule. I believe that much better results are 
obtained through a process of evolution, by starting with 
some definite trait and developing it through a series of 
crosses. To be sure, this process is much slower and requires 
patience and study, but it is scientific and, to me, the results 
obtained are far more gratifying than those obtained by what 
must be considered as more or less chance crosses. 
In my work I seldom figure a cross in advance: rather, each 
cross is to me an individual problem, to be met and solved as 
it arises. Having selected any given variety for the seed 
parent, I set about to find the pollen parent that seems to 
