282 NOTES ON SARCOMENIA MINIATA AG. pericentral tube. The stem is of a narrow elliptic form, since, seen on a cross-section, there are three cells lying one behind the other in one direction, and five cells in the other. The pericentral cells have not the same morphological value, for a cross-section shows clearly that the diameter of the lateral tubes from front to back is nearly as broad as the diameter of the three middle cells taken together (fig. 2). Fig. 1 is a diagrammatic drawing of the top of a branch, and shows the way in which the tubes and cortical cells are developed. This process is exactly the same as that described by Prof. Cramer* for the first divisions of the genus Caloglossa ; the structure of Caloglossa, however, being more complicated than that of Sarco- menia, the subsequent changes that take place in the segments are different. The initial cell of Sarcumenia divides first by a horizontal wall ; the segment thus cut off shows first one and then a second vertical partition, dividing the segment in three equal portions. The exact moment when the tangential walls appear, that cut out the front and back pericentral tube, could not be ascertained, but all these divisions take place very quickly one after another. The lateral tubes are afterwards partitioned by two oblique walls extending from the middle of the upper or the lower wall to the middle of the outer wall, and the segments thus cut off form the row of cortical cells. I was anxious to know whether some of the adaptations in Caloglossa to prevent the decay of the stichidia after the departure of the tetraspores were to be found also in Sarcomenia; my supposition was confirmed for the branches destined to become stichidia ; the lateral tubes, after having given off the cortical cells, divide again by a vertical wall into two equal cells. The back one — by which name for clearness' sake I dis- tinguish the cell that will bear no tetraspores — undergoes no further changes. The front one is cut in two by a horizontal wall ; the upper cell resulting from this partition becomes the mother-cell of the tetraspores ; the lower one is a kind of stalk-cell to the sporangium, and may also serve to consolidate the stichidia (figs. 3, 4, 5). During this process each cortical cell divides again by an arcuate wall into a small upper and a bigger lower cell, this last one bearing some resemblance to a crescent with rounded horns. After the maturity and escape of the tetraspores, these horns push forth to the empty sporangia from where the tetraspores have fallen out by a slit in the membrane, and dividing themselves give rise to cortical cells that take the place of the sporangia. This has a double advantage for the plant: first, the vegetative life of the stichidia will be lengthened, for these new cortical cells will protect the stichidia from being torn off by the current, which might easily happen after the tetraspores have left the sporan<;ia ; and secondly, it must not be forgotten that the branches of Sarcomenia miniata have all an acropetal growth. The tetraspores at the base of the stichidia are long ripe, have in fact often left the sporangia, before Two of these cortical cells have the same length as one cell of the
- Cramer, C, Ueher Caloglossa Leprieurii Harv., J. G. Ag. Tab. ii. figs. 1 & 2.
