130 ANIMALS KILLED Iv, 
did not revive. To afcertain whether the fa& 
was merely accidental, I fpread a portion of the 
fand on a glafs flider, and waited until the nu- 
merous reanimated wheelers became dry in or- 
der to wet the fand anew. The fand was care- 
lefsly fcattered on the glafs fo as to bea thin 
covering on fome parts and on others in very 
fmall quantity: here the animals did not revive, 
but all that were in thofe parts with abundance 
of fand, revived. A difference fo remarkable 
made me fufpect that, the ordinary dwelling of _ 
thefe animals being in fand, a certain portion 
mutt be prefent to enable them, to pafs from death 
to life. 
However, to acknowledge the truth, though I 
did not at firft adopt this conjecture, I could not 
diveft my mind of it, as it feemed to be confirmed 
by facts; befides, on recolleGting the experiments 
before made on wheel animals, they had certain- 
ly always been in fand. ‘To corroborate or con- 
fute the fact, it was only neceflary to repeat the 
laft experiment; for, if thofe that revived had 
been mixed with fand at the moment of refur- 
rection, or, on the contrary, if thofe without 
fand did not revive, it would be complete de- 
monttration of what had been the caufe, namely, 
that the prefence of fand was effential to their 
refurrection. On repeating the experiment, it 
conftantly followed that the animalcula never re- 
covered life unlefs in places where there was a 
quantity 
