4 8 
The Ohio Naturalist. 
[Vol. VII, No. 3, 
Explanation of Plate IV. 
The figures were drawn with the aid of an Abbe camera 
lucida. Figure 1 was drawn with the Zeiss 18 ocular and Leitz 
1-16 objective; the others with Zeiss 18 ocular and Zeiss 8.0 ob¬ 
jective. 
Fig. 1 . A single bivalent chromosome from the reduction cell in the 
ovule of Lilium philadelphicum. 
Fig. 2. Nucleus from two-celled sac of L. philadelphicum showing 
expansion of the cytoplasm. 
Fig. 3. Microsporoeytes of Marsilea quadrifolia showing centrifugal 
arrangement of the chromatin in synizesis. 
Fig. 4. Microsporocyte tissue of Marsilea quadrifolia showing sym¬ 
metrical synizesis. 
Figs. 5 and 6. Cell rows from opposite sides of sections of Allium 
cepa showing the nature of artificial contraction. 
Figs. 7 and 8. Microsporoeytes of Sagittaria latifolia with nucleoli 
on one side of the contracted chromatin. 
Fig. 9. Microsporocyte of Erythronium americanum showing cen¬ 
tral contraction of the chromatin around the nucleolus. 
Figs. 10 and 11. Microsporoeytes of Lilium tigrinum showing the in¬ 
dependent contraction of the chromatin, the nucleoli being entirely dis¬ 
tinct in the nuclear cavity. 
WEATHER AND CROP YIELD* 
J. Warren Smith, Section Director, Weather Bureau. 
I am convinced that the yield of most of our staple crops is 
affected greatly by favorable or unfavorable weather conditions 
during a comparatively short period of their growth. 
Also that the yield can be very closely approximated at the 
close of this critical period by an exhaustive study of the weather 
conditions and crop yields of previous years. 
The United States Weather Bureau is accumulating a vast 
amount of weather data that are being splendidly tabulated for 
use in studies of this character and considerable attention is 
being given to the subject. 
In 1902 Professor W. D. Gibbs and myself charted some very 
interesting curves showing the relation between the yield of 
corn in the eight great corn producing states in this country, 
and the rainfall during June, July and August. 
The period covered was from 1888 to 1902, and the states 
considered were Ohio, Indiana, Illinois, Iowa, Nebraska, Kansas, 
Missouri, and Kentucky. 
The result showed that the yield was very little affected by 
the rainfall in August and not greatly affected by the rainfall 
during June. The curves, showed, however, that the rainfall in 
July controls the corn yield to a marked extent. 
* Read at the meeting of the Ohio State Academy of Science. 
