168 GRIFFITHS: A NOVEL SEED PLANTER 
very different appearance from that on the deserts of Arizona. 
Instead of being sunken ina small pit they simply form the center 
of little wads of sand which are firmly attached to them by the 
adhesiveness of the mucilage. The seeds were slightly sunken but 
there was no sign of pits in the sand. 
When the seed sown on the sand of a roadbed had dried, some 
of it had exactly the same appearance as that upon the clean build- 
ing sand, while the remainder presented exactly the same appear- 
ance as that upon the hard limy soils in the vicinity of Tucson, 
Arizona. 
That sown on the hard smooth surface, of reasonably stiff 
clayey soil differed in no way when dry from that described above 
for its natural habitat. The pits with the seed seated in them were 
as perfectly formed as one could wish to see them. 
Bringing the phenomena of these different experiments together, 
it appears to me that we can formulate an explanation of the action 
of the mucilage in the sinking of the seed and the forming of the 
pit into which it gravitates, or rather is both pushed and pulled. 
When the mucilaginous coating is completely distended it measures 
about 5x 7 mm. in its greatest horizontal diameters. The outer 
portion of the coating, especially, flattens out wonderfully and pene- 
trates the substratum as shown by the experiments on paper 
When the process of desiccation begins, the soil being porous will 
give up its moisture more readily both to the underlying strata, 
which are almost invariably dry in Arizona, as well as to the 
atmosphere than the mucilaginous seed coat. 
This condition together with the fact that the outer layers ee 
moisture much more rapidly than those close to the seed results in 
drying and fastening the outer edge which has flattened out upon the 
soil and also sunken into it, as well as the lower surface, firmly to 
the soil particles. As soon as this occurs a tension is set up which 
would naturally be toward the center—that is the outer contract- 
ing layers would crowd the inner ones—but having meee 
anchored to the soil particles and rendered rigid by drying, 4" 
having only a loose connection with the undried inner gam “ 
pull is outward in all directions. The underside of the mucl pet : 
nous particle in contact with the soil has also lost moisture, : 
here to a much less degree than the outer edges in better contac 
? 
