6og 
and their Mycorhiza. 
branched lateral roots at intervals along the main ones. Whether this 
frequent branching is a direct outcome of the presence of the fungus, which 
would exert a stimulus on the tissues, or not, is a point that cannot, of course, 
be decided. From analogy with recent plants, this seems highly probable. 
Recent work on the physiology of mycorhizic plants seems to show 
that the relationship of the fungus to the host is a very vital one. And 
whether it is one in which the host benefits by obtaining a more abundant 
supply of mineral salts, from a soil rich in humus, which it is unable to do 
efficiently without the aid of the fungus, according to Stahl (17), or whether 
the endophyte has the property of fixing atmospheric nitrogen, so giving its 
host a larger quantity of nitrates than it could otherwise obtain (Nobbeand 
Hiltner, 9, and Hiltner, 4) the general result seems fairly established. 
Myrica (when it grows in swampy soil), Alnus and Elaeagnus all 
produce short fleshy roots caused by either a fungus or a bacterium and 
inhabited by it, and the benefit accruing to the spermophyte in at least one 
of these cases has been strikingly demonstrated (Moller, 8). It is not 
unreasonable to suppose in that case that the rootlets in Amyelon , which in 
many respects seem to have had a similar structure to recent mycorhizal 
roots, may have had a similar function. 
Stopes and Watson (18), in their paper on ‘ Coal Balls have given 
cogent reasons to prove that these concretions were formed in situ by 
gradual petrifaction of vegetation at the bottom of a brackish swamp. The 
conditions of life in such a locality are exactly of the kind that would lead 
one to expect the production of root-tubercles on some of the plants grow- 
ing there. The soil would be deficient in nitrates, while it would be rich in 
humic acid. Mycorhiza are frequently associated with plants liable to 
suffer from drought, or growing in a soil abundant in humus. These are the 
conditions that we might expect in a saline marsh, the drought being 
physiological rather than physical (Schimper, 12). Indications are not 
lacking that the leaves of some species of Cordaites (C. crassus and angulo - 
striatus ) were possessed of xerophyllous characters (Renault and Scott). 
It has been pointed out that these rootlets are found mingled with, 
and growing amongst, a mass of debris of other plants of all kinds. 
While the Amyelon roots are frequently well preserved, the fragments of 
vegetation around are often much decayed. It is quite easy to see that 
such would be the appearance in a petrifaction formed in situ , if the 
Cordaites roots were growing there, the other plants being mere debris. 
These, indeed, are very much the conditions in which Myrica Gale 
is to be found growing around the Cheshire meres to-day, only the water 
is fresh and not salt, as it probably was in a Coal Measure swamp. An 
examination of these rootlets has led me to the conclusion that they were 
probably borne by plants adapted to such conditions as would exist in the 
kind of marsh that Stopes and Watson describe. 
