1894. ] Mutualistic Symbiosis. 27 
rhizobia, bacteria, hyphal fungi and various species of alge, 
the discussion of which will be taken up below. The second 
layer is the dermatogenic layer of tangentially elongated rect- 
angular thin walled cells. The third layer is the subdermal 
parenchyma of large rounded cells with intercellular spaces 
and air conducting passages. Considerable starch is present 
and often oil globules are found toward the inner side where 
the most rhizobia and bacteria are also found. The fourth 
layer is of the most importance. It consists of two opposite 
rows of palisade cells. This layer is only present in tubercles 
bearing algz and is formed from radially elongating paren- 
chyma cells beginning near the apical area of the tubercle and 
extending near the point of separation between tubercle and 
rootlet. The two rows of palisade cells are separate or only 
loosely connected in the middle. The cells are thin walled, 
about two or three times as long as broad, with large nuclei 
suspended in a granular cytoplasm. The large intercellular 
spaces are entirely filled with alge (Wostoc sp. ?). Besides the 
granular cytoplasm, the cells contain starch, amyloplastids, 
sometimes oil globules, and a waxy body near the base. The 
fifth layer or parenchyma proper resembles the subdermal 
parenchyma. The cells contain much starch. Numerous 
cells entirely filled with a waxy substance are present. The 
sixth layer is the vascular system sheath consisting of modi- 
fied parenchyma cells of several layers thickness. The vas- 
cular system need not be described as it is the same as that 
of the ordinary root. 
A longitudinal section shows the presence of a rudimentary 
root cap consisting of elongated loosly connected cells cover- 
ing more or less perfectly the rounded end of the tubercle. 
The apical area consists of small prismatic closely united 
meristem cells. The palisade cells do not extend quite to 
the apex. (See plate 111, fig. 2.) 
The exact cause of the development of these tubercles I am 
unable to state. That there is excessive metabolism is very 
evident from their appearance and the large amount of albu- 
minous substances present. That the infecting alge are not 
the cause of their development is shown by the fact that tu- 
bercles exist without the alga-bearing palisade layer. On 
Making a comparative study of Cycas roots and tubercles I 
found the following differences: The dermal and apical area 
of tubercles contained more rhizobia and bacteria than 
