890 
Journal of Agricultural Research 
Vol. XXIII, No. II 
perature was high- or low. By reducing the moisture supply after a 
foliage shoot some 10 to 12 cm. high had developed as a result of an 
abundant supply of moisture, further upward growth was stopped and 
numerous small aerial tubers soon formed. The picture thus presented 
( 2 3> Table IV , p. 114) strongly calls to mind the general contour of a 
miniature shrub, defoliated but bearing numerous fruits. With moist 
soil but dry atmosphere the shoots which developed soon assumed a 
horizontal direction and numerous basal branches appeared. In germi¬ 
nation tests in diffuse light a moist soil gave relatively large leaves and 
long intemodes in the presence of moist air and short, thick intemodes, 
mere scales rather than true leaves, and basal stolons in the presence of 
dry air. Tubers set upright in plates, without use of any soil, showed 
decreased elongation of the apical shoot with increased light intensity. 
Under these conditions light inhibited root formation. 
In contrast with Voch ting’s work on tuber formation, Bernard, in a 
lengthy article ( 1 ) dealing largely with the life history of the orchids, 
reached the conclusion that in many species a period of active differ¬ 
entiation is followed by a definite period of tuberization, while in other 
species tuberization begins in the earliest stages of development. In all 
cases Bernard found a close correlation between tuberization and infection 
with certain endophytic fungi and concluded that these parasites are 
the cause of tuber formation. This fungus theory of tuberization was 
extended to include the potato. It had been previously shown by 
Laurent (12), however, that shoots from potato tubers, which had been 
grown in darkness and were free from starch, developed tubercles in 
darkness when placed in saccharose solutions of proper concentration 
(10 to 20 per cent). Having subsequently confirmed these results 
Bernard (2) was forced to modify his views as to infection by fungi being 
the sole cause of tuberization. Molliard (r6) showed that onion (.Allium 
cepa L.) grown in a sterilized nutrient solution containing glucose is 
capable of forming a bulb and later (17) showed, also, that radish and 
certain other species form a starchy, thickened root under similar con¬ 
ditions. Thus, while it may be possible that in special cases tuberization 
is due to the invasion of fungi, it seems certain that parasitism or sym¬ 
biosis is not the usual cause of the formation of tubers and other storage 
organs. Magrou, in a recent extended discussion (15), however, seeks to 
establish the importance of Bernard’s fungal theory of tuberization. 
“Sets” of the common Silverskin variety of onion were grown under 
different day lengths, 15 to 20 individuals being grown in each lot. In 
the first test the sets were planted in the greenhouse May 19. One lot 
was exposed to the natural*day length while the second lot received in 
addition electric illumination of 3 to 5 foot-candles from sunset till 
midnight in the manner previously described. The summer temperature 
in the greenhouse was high, averaging io° to 20° F. above the outside 
air temperature. The behavior of the plants both with and without 
artificial illumination was much the same as that of the outdoor controls 
except that the size of both tops and bulb was considerably reduced, as 
shown in Plate 9, A. The controls exposed to the full day length of summer 
developed in the usual manner. First blossoms opened July 14, large 
bulbs were formed, the tops died in due course, and there was no splitting 
into new individuals. The resting bulbs resumed vegetative develop¬ 
ment in September. A series exposed to 13 hours of light daily out of 
doors did not flower and the tops remained green throughout the summer. 
