aquilinum , with some Notes on Marsilia quadrifolict , &c, 579 
alteration of the substance of the middle lamella between the callus plugs 
takes place. This change in the middle lamella causes it to have the 
property of staining with protoplasmic dyes, and after treatment with 
safranin the altered places are very conspicuous as red dots (PI. LIV, Figs. 5 
and 6). The threads stain more darkly at this stage, but are still very 
slender, so that they are not very conspicuous ; before treatment with 
London blue they can generally be traced fairly easily, but after the callus 
has been stained, the deep blue colour is often sufficient to obliterate the 
outline of the slime strings altogether (PI. LIV, Fig. 4). Each slime string 
possesses its own distinct pair of callus basins (PI. LIV, Fig. 9), though when 
two slime strings are sufficiently near together their callus basins may touch 
one another, and give the appearance of more than one string contained in 
a single mass of callus (PI. LIV, Fig. 6). Soon after this stage has been 
reached, it is quite apparent that the arrangement of the refringent granules 
bears a very definite relation to that of the slime strings and callus basins 
(PI. LIV, Fig. 12). The granules are larger than in the earlier stages, and their 
appearance suggests that they are more of the nature of mucilaginous 
droplets than of granules. A refringent granule is attached to each end of the 
slime string (PI. LIV, Figs. 12 and 17), and in a section of a sieve plate each 
slime string has the appearance of a hatpin with a knob at each end. 
Whether, however, the callus is of a sticky nature and causes the granules 
to adhere, or whether they are actually connected with the slime strings, 
could not be determined. In a surface view of a sieve plate the refringent 
granules, which stain darkly with safranin, make the callus basins invisible 
(PI. LIV, Figs. 13 and 14), though on focusing carefully the granules take 
a mauve tinge, due to the blue of the callus below them ; in sections of this 
stage, stained in London blue only, there can be seen successively, on focusing 
downwards, the clear outline of the upper refringent granule, then a blue 
spot (the callus rod), and lastly the outline of the lower refringent granule. 
In sections stained with both London blue and safranin the callus can only 
be seen here and there in surface view as a blue spot, where the refringent 
granule above it has been dislodged (PI. LIV, Figs. 15 and 16) ; in such cases 
the median nodule can be focused as a red spot in the midst of the blue 
(PL LIV, Fig. 1 5). The median nodule is also visible as a dark spot on 
focusing down through the granules (PI. LIV, Fig. 15). The granules seem 
to grow larger after their attachment to the ends of the slime strings becomes 
apparent (PL LIV, Fig. 16). When this stage becomes visible the callus 
basins begin to be dissolved away (PL LIV, Figs. 18 and 19), possibly through 
the action of the granules, which sink down into the cavity left by the dis- 
appearance of the callus. The median nodule, also, next becomes entirely dis- 
solved away, so that there is now a thick slime string (PL LIV, Figs. 18-26) 
stretching uninterruptedly across the pit-closing membrane, and derived 
apparently, in part at least, from the refringent granules or mucilaginous 
