22 
Proceedings of the Royal Society of Edinburgh. [Sess. 
On examination the deposit was found to consist of iron-bacteria which 
have not hitherto been described. Although I have examined samples of 
iron- water from several places, this is the only place in which I have found 
this particular organism. The question of its distribution must be left to 
another paper. I propose to call the genus Spirophyllum , and this 
particular species Spirophyllum ferrugineum. 
General Structure. 
A typical example of this plant is seen in fig. 1. The body of the cell 
is flattened like a leaf, and spirally twisted. In this figure two turns are 
shown ; but there is great variety in this respect, for, while there ma}^ be 
only half a turn in some, in others there may be fifteen or more complete 
turns. By a complete turn is meant one in which, as a result of the turn, 
the lower surface has become the upper surface. The name “ Spirophyllum ” 
is meant to indicate the above facts, viz., a spirally wound, flattened 
structure. 
With regard to the dimensions, it is necessary to state them as regards 
length, width, and thickness. In addition, in this case we must also state 
the length of a turn or a twist in the same direction. The width is least 
immediately after germination from the conidia, e.g. figs. 16, 18, in which 
the width is 1 jx. In fig. 2, which shows a portion of a spiral which 
had fourteen turns, the width is nearly 6 jx. 
Between these two dimensions all sizes are seen, showing a perfect 
gradation, leaving us in no doubt as to the inclusion of all these forms 
under one species. The maximum length, of course, cannot be stated. In 
fig. 3 we see what is very probably the first stage in the germination, in 
which the length is not more than about 2 /x. On the other hand, fig. 2 
in its entirety would reach up to 180-200 /x. I have no doubt that still 
longer threads exist, because, as is explained later, this organism probably 
relies on the germination of conidiospores, and not on vegetative division, 
for increase in numbers. The thickness of the bands is not easy to 
determine accurately, because, though the threads are often seen with the 
edges facing the observer, as must be the case in a spirally twisted organism, 
yet the edges are thicker than the middle portions. The edges attain a 
thickness extending to approximately \ /a ; the middle portion can be roughly 
estimated at jx. 
The closeness of the spiral turns finds its highest expression in fig. 4, 
which is, however, extremely exceptional. In this case the length of a 
twist is very little greater than the width. The average is seen in fig. 1, 
