30 
Proceedings of the Royal Society of Edinburgh. [Sess. 
activities are rendered possible. The iron-bacteria are thus supposed to 
respire, and so obtain their energy much in the same way as the sulphur- 
bacteria had been previously proved by him to obtain their energy. In the 
one case the free sulphur is oxidised to the sulphate, in the other the ferrous 
compound is oxidised to the ferric form. This conclusion was based upon 
the following data : — 
1. Deposition of iron occurs only in water containing iron in the ferrous 
condition. 
2. For the growth of these bacteria ferrous carbonate is absolutely 
necessary. 
3. Only living threads possess an iron deposit. 
Hence the conclusion that the presence of a large quantity of iron is 
absolutely necessary to the existence of the organism. The first two of 
the above data were never proved by him. His publication was meant 
as a preliminary notice, but, though nineteen years have elapsed, no further 
publications from his pen on this subject have, to my knowledge, appeared. 
It is also remarkable that in his prescription for the cultivation of these 
bacteria he recommended that freshly precipitated ferric hydroxide be added 
to the culture medium, presumably to supply the necessary energy. If 
respiration be dependent on the oxidation of a ferrous to a ferric compound, 
one would have expected that a ferrous compound should have been added. 
The improbability of this hypothesis — for he did not carry it beyond this 
stage — is obvious when it is remembered that when a solution of ferrous 
carbonate is made, it changes into the hydroxide almost immediately, and 
very special precautions have to be taken to prevent its doing so, by care- 
fully eliminating the atmospheric oxygen. An important work on this 
subject was published by Molisch (4), who was able to cultivate bacteria in 
solutions in which not a particle of iron was present. He rightly explained 
the reason why living threads are able to store iron, but dead threads were 
not able to, by stating that living threads were surrounded by a mucilaginous 
layer, but that dead threads were devoid of it, and, as it is this layer which 
retains the iron, this substance is obviously absent from dead threads. I 
have verified this fact in the case of Leptothrix ochracea, in which obser- 
vation on this point is rendered easy by its sharply defined membrane. 
Again, if it were true that iron is absorbed into the system, we should be 
able to detect it inside the cell. Winogradsky does not mention that he 
was able to do this, and neither Molisch nor myself have succeeded. The 
deposit can very easily be cleared away by adding a little very dilute 
hydrochloric acid under the coverslip to a microscopic preparation. In the 
