28 BULLETIN 1195, U. 6. DEPARTMENT OF AGRICULTURE. 
In physiological incompatibility, well exemplified in the lilies, fruit and seed 
production depends on the stimulus of compatible fertilizations, and the bulb 
or tuber forming plant is consequently found to be fully capable of developing seeds 
when the proper and compatible pollinations are made. 
(2) Experimental proof of a direct and simple competition for food between tubers 
or bulbs and the forming seeds is lacking. The older statements frequently quoted 
that the flowering branches of such plants as Lilium candidum and Hemerocallis fulva 
will produce fruit and seeds only when they are cut and placed in water and that 
the removal of aerial bulblets from such plants as Lilium tigrinum and Ficaria ficaria 
will lead to seed formation are incorrect. This matter has been tested carefully in 
experiments at the New York Botanical Garden with results showing that when pods 
are produced in such cases the plants are to some degree self-compatible and able to 
produce pods on branches left attached, provided the same care is taken in making 
the pollinations. 
Experiments with the potato which attempt to test for correlative sterility have 
all given negative results, at least since the favorable results reported by Knight (14, 
p. 58). In an experiment of this sort East (4, p. 432-433) prevented the formation 
of tubers in plants of the three varieties Rural New Yorker No. 2, Green Mountain, 
and Irish Lemon. In comparing treated with untreated plants he found no increase 
of apparently good pollen, and to pollinations of 50 plants of each he obtained 6, 9, 
and 2 fruits, respectively, for the treated, and 4, 8, and 4 fruits for the untreated. 
The marked fruitlessness of these varieties to self-pollination was not changed by 
this treatment. 
It would seem that a critical experimental test of a direct correlative sterility in 
potatoes could be made by grafting branches on species of Solanum that do not form 
tubers. A few years ago an extensive experiment of this sort was conducted by 
Prof. C. H. Myers and an assistant, W. I. Fisher, at the Agricultural Experiment 
Station of Cornell University. The results have not been published, but a copy of 
the report has been kindly furnished by Professor Myers to the writers. Branche's 
of 16 varieties were grafted on various species of Solanum which do not form tubers. 
The varieties thus used were Beauty of Hebron, Blue Victor, Carman No. 3, Early 
Ohio, Early Rose, Green Mountain, King Edward VII, Moravia, Pat's Choice, Phoe- 
bus, Rural New Yorker, Sir Walter Raleigh, State of Maine, Vermont Gold Coin, 
Wohltmann, and an unknown sort. Most of the grafts were made on Solanum ciliatum, 
S. sysimbrifolium, S. miniatum, and S. nigrum. In all, 200 successful grafts were 
made, and the plants were grown to good maturity out of doors in a garden. Many 
careful hand pollinations (evidently self) were made on flowers, but only one seed 
ball was obtained from the whole lot. In these grafted plants there was no tuber 
formation to draw food from the aerial parts. The results are positive in indicating 
that fruitfulness in these plants is not increased when the qualitative competition 
for food between vegetative organs and organs of seed production is removed. The 
sterilities to own pollen, here due in large part to impotence of pollen, or also to ab- 
scission were not removed by such experimental methods. There was no indication 
in the results that the nature of the flowers and of their sex organs was influenced 
or changed. 
(3) In a careful study of the osmotic pressures in potato plants with reference to 
the distribution of food at various stages of growth Lutman (15) has shown that while 
a superior osmotic pressure seems to be necessary for the formation and growth of 
new sprouts the growth of berries (fruits) and of tubers can not be thus accounted 
for, since the osmotic pressures in these organs are the least of any in the plant. Lut- 
man points out that there appears to be a pumping action of the sieve tubes, but 
that a study of the comparative osmotic pressures does not reveal the manner in 
which they may induce the observed flow of food materials even to the tubers. 
(4) For the cultivated varieties of the potato the evidence is clear that when con- 
ditions favor blooming the use of viable pollen gives fruit in abundance. Fruit is 
often abundantly found in the field, especially on varieties having considerable good 
pollen. Furthermore, a study of the comparative yields of tubers and seed balls for 
plants of such a variety (Lookout Mountain or McCormick ) by Newman and Leonian 
(17, p. 14) has shown that 'in most cases the heaviest tuber production, the richest 
vegetative growth, and the largest seed productions go hand in hand." There are, 
however, various conditions influencing the growth of plants, the yield of tubers, 
and the production of seed balls. Under certain of these conditions there may be 
internal regulation of development that directly or indirectly determines whether 
the success or failure of the stimulation of fertilization is to be effective. 
The main point of the evidence noted above regarding the setting of 
seed is that under favorable conditions for blooming, potatoes are well 
