A EEMAEKABLE MAEINE OEGAKISM. 445 



which does not stain with any of the above reagents, are met 

 with, as at I, I in fig. 5. On treatment with sulphuric acid these 

 granular layers dissolve with brisfi; effervescence and forma- 

 tion of needle-shaped crystals in the neighbourhood: we may 

 therefore conclude with some degree of confidence that they 

 consist of carbonate of lime deposited by the organism. They 

 are not to be confounded with tlie foreign particles of sand which 

 occur imbedded in the deeper layers as at g, g in fig. 6. 



Careful examination of the transparent ground-substance 

 between the flocculent, deeply-staining material shows that the 

 former is not by any means structureless as it appears at first 

 sight, but contains, or perhaps one should even say consists of, 

 innumerable exceedingly slender unbranched threads, lying close 

 together, and generally, but by no means always, arranged in a 

 radial manner, more or less at right angles to the successive 

 surfaces of growth. These threads are most strikingly brought 

 into view by mounting in chlor-zinc-iodide a thin section 

 previously soaked in a solution of iodine in potassium iodide 

 (PL 27. fig. 7). They then assume a beautiful blue or violet 

 colour, which is also produced by treatment with iodine and 

 sulphuric acid and indicates that they consist, at any rate largely, 

 of cellulose. The threads are only about O'OOIS mm. in diameter, 

 and further examination shows them to be the cellulose sheaths 

 of chains of short, rod-like bacteria. In the deeper layers of 

 the organism nearly all the sheaths are empty, but a few may 

 still contain the chains of bacteria to which they owe their 

 origin (PI. 37. fig. 8). 



The bacteria themselves are stained yellow or brown by the 

 action of the iodine in the chlor-zinc-iodide method, and are then 

 clearly seen to consist of short rods, somewhat less in diameter 

 than the containing sheath, only about 0-00135 mm., and two 

 or three times as long as they are broad. The size of the rods, 

 however, varies a good deal, as will be seen by reference to 

 fig. 8. As we approach the surface we find more and more of 

 t he rods, until in the youngest layer of the colony we have a 

 dense and almost solid mass of more or less felted chains of rods. 

 Thus fig. 9 shows a small portion of a thin vertical section cut 

 by the paraffin method and stained, after drying on the slide, 

 with carbol-fuchsin, followed by iodine. It will be noticed that 

 the chains of rods are very distinct, having taken the stain 

 strongly, but the cellulose sheaths are not visible. This figure 



