282 BOTANICAL GAZETTE [OCTOBER 
microscopical examination of the soil shows that it consists of 
irregular, very angular grains of many different colors and sizes. 
Very marked and characteristic, in every one of the very many 
samples of the soil examined, is the occurrence of fragments of 
sponge spicules, some of the more marked forms of which are 
shown in jig. 6. While 
WA these spicules do not 
| form any considerable 
\ part of any particular 
\ sample, certainly not I 
per cent., nevertheless 
their aggregate quantity 
in the enormous deposit 
5 ‘smm- of marsh mud is very 
Bees type forms of sponge spicules great, supposing, as the 
samples indicate, they 
are distributed everywhere through it. I do notimagine, however, 
that their presence has any significance whatever in the ecology 
of the vegetation.** Numerous diatoms of several forms are also 
present in the mud. 
The mechanical composition of a soil is important chiefly 
because of its relation to the supply and circulation of air and ~ 
water through it. The finer a soil is, other things being equal, 
the better will it hold water in the hygroscopic state, and hence 
the better it is for the constancy of water supply to the vegeta- 
tion. Buton the other hand the finer it is the less air will it hold 
and allow to circulate, and air (7. ¢., the oxygen) necessary for 
the respiration of roots is well-nigh as essential a constituent of 
the soil as water. The soil of the marshes, being much finer than 
the average, is better than the average for holding and delivering 
76T think they are without doubt spicules, eb they are very say mee ady Rare 
are insoluble in the ordinary acids, and hence are probably siliceous. Very milar 
forms are figured in the Challenger Report 11: g/. 3, fig. 6, though I have ae seen 
= branched forms. Two sources of s supply are imaginable; living sponges in the 
and the fossils in the Permo- Carboniferous rocks. But the water must be toomuddy 
floated up from the decay of sponge bodies. Dr. G. F. Matthew tells me fossil sponges _ 
are not known from the Permo-Carboniferous rocks. 
