The quantitative estimation of the micro-organisms present in air is a problem which has attracted the attention of many experimenters since it received its first impetus at the hands of Pasteur twenty-six years ago. The method originally employed by Pasteur was of a very simple character, and consisted in preparing vacuous flasks of definite (1/4 litre) capacity, each containing a small quantity of nutritive liquid : the air in the flask was removed by boiling the liquid, and the open extremity was then sealed with the blowpipe. Such sealed flasks could then be preserved for an indefinite period of time without undergoing change.