Mitscherlich was also convinced of the vegetable character of yeast, and showed [1841] that when yeast was placed in a glass tube closed by parchment and plunged into sugar solution, the sugar entered the glass tube and was there fermented, but was not fermented outside the tube. He regarded this as a proof that fermentation only occurred at the surface of the yeast cells, and explained the process by contact action in the sense of the catalytic action of Berzelius, rather than by Liebig's transference of molecular instability. Similar results were obtained with an animal membrane by Helmholtz [1843], who also expressed his conviction that yeast was a vegetable organism. In 1854 Schröder and von Dusch [1854, 1859, 1861] strongly reinforced the evidence in favour of this view by succeeding in preventing the putrefaction and fermentation of many boiled organic liquids by the simple process of filtering all air which had access to them through cotton-wool. These experiments, which were continued until 1861, led to the conclusion that the spontaneous alcoholic fermentation of liquids was due to living germs carried by the air, and that when the air was passed through the cotton-wool these germs were held back. At the middle of the nineteenth century opinions with regard to alcoholic fermentation, notwithstanding all that had been done, were still divided. On the one hand Liebig's theory of fermentation was widely held and taught. Gerhardt, for example, as late as 1856 in the article on fermentation in his treatise on organic chemistry [1856], gives entire support to Liebig's views, and his treatment of the matter affords an interesting glimpse of the arguments which were then held to be decisive. The grounds on which he rejects the conclusions of Schwann and the other investigators who shared the belief in the vegetable nature of yeast are that, although in some cases animal and vegetable matter and infusions can be preserved from change by the methods described by these authors, in others they cannot, a striking case being that of milk, which even after being boiled becomes sour even in filtered air, and this without showing any trace of living organisms. The action of heat, sulphuric acid, and filtration on the air is to remove, or destroy, not living organisms but particles of decomposing matter, that is to say, ferments which would add their activity to that of the oxygen of the air. Moreover, many ferments, as for example diastase, act without [p011] producing any insoluble deposit whatever which can be regarded as an organism. "Evidemment," he concludes, "la théorie de M. Liebig explique seule tous les phénomènes de la manière la plus complète et la plus logique; c'est à elle que tous les bons esprits ne peuvent manquer de se rallier."