OF ARTS AND SCIENCES. 33 



in the activity of the hyphae which lie immediately below the epi- 

 dermis. Certain of these hyphae are seen to send out processes 

 which grow out towards the surface of the leaf. These outgrowths 

 push their way between the epidermal cells, but do not at this early 

 stage enter them, soon becoming, however, multicellular by the 

 formation of frequent cross septa (Figs. 1, 2). At first the cells 

 which surround the developing spermogouium arc but little affected, 

 the rapidly growing hyphae merely splitting the wall at the middle 

 lamella. The space between the cell walls, thus separated and 

 forced apart, affords a cavity in which the hyphae may grow. The 

 hyphal threads now increase very rapidly in number, both by di- 

 vision and by the sending up of new outgrowths from the hyphae 

 beneath, so that the parted walls of the adjoining epidermal cells 

 are very much pushed outwards to make room for the developing 

 spermogouium (Fig. 3). At the same time, these hyphae, which 

 were at first very irregularly arranged, show a tendency to place 

 themselves in parallel rows, and their contents begin to assume a 

 more granular appearance than those of the ordinary vegetative 

 hyphae. Both of these conditions become more marked as the 

 spermogonium matures. 



As the spermogonium continues to increase in size, it exerts more 

 and more pressure on the walls of the cells which confine it, and 

 having distended them outwards as far as possible finally ruptures 

 them (Fig. 4, a). The rupture takes place from the atrophy of the 

 walls, which have by this time become very much weakened by 

 pressure. The hyphae are now free to grow into, and eventually fill, 

 the larger space afforded by the union of the cavities of the sur- 

 rounding cells. The remnants of the ruptured walls arc apparently 

 soon absorbed or hidden by the growing hyphae, for in stages of the 

 spermogonium not a great deal more advanced than this they are 

 very hard to distinguish. It is usually a long time, however, be- 

 fore all traces of the separating wall are lost, as will be nut iced in 

 Figure 4, b, for even in this much later stage indications of the 

 wall may still be seen in a slight protuberance, which extends 

 downwards into the cavity of the spermogonium. The tune in the 

 development of the spermogonium at which the rupture takes place 

 is not at all constant; in some cases where the mass of hyphffl was 

 very small, it was seen to have already broken through the walls. 

 After the hyphae have made their way into the larger cavity, their 

 parallel arrangement becomes more conspicuous than before, and as 

 the hyphae continue to multiply in number and increase in si 



VOL. XXVIII. (n. S. XX.) 3 



