Bacterial Flora of the Semi -Desert Region of New Mexico. 231 



Aug. 8 — Plates well developed. Early count is necessary. 

 PI. r — Contaminated by melted paraffin. 

 PI. 2 — 42 bacteria and 2 molds. 

 PI. 3 — 58 bacteria and 3 molds. 

 Average per 10 min., 8.3 bacteria. 

 This is lower than in previous experiments, probably due 

 to quiet atmosphere and to early counting of plates. The 

 species are A 1( A 2 , A,, A 4 , A 13 and A u . The extent of this 

 flora was somewhat of a surprise to the writer, for only one 

 or two species had been expected. A n and A u were most 

 numerous and both were spore-bearing bacilli. 



The three experiments just recorded would seem to leave 

 no room for doubt regarding the fact that bacteria are present 

 not only above 5,000 ft. of altitude, but that they may be 

 found even on the mountain peaks at an elevation of 10,000 ft. 

 at least in the summer season. This fact is in perfect har- 

 mony with the general law that wherever vegetation and 

 animals may exist there the bacteria will likewise be found. 



Quantitative Data. 



In addition to the qualitative work that has been recorded, 

 a number of determinations were also made quantitatively. 

 For this purpose two methods were resorted to. In experi- 

 ments 33-38, an 18-liter oil can, filled with water, was used as 

 an aspirator to remove the air from a large (500 cc) Erlen- 

 meyer flask. This flask was fitted with a rubber stopper 

 through which entrance and exit tubes passed. These tubes 

 were plugged with cotton, and a quantity of sterile gelatin 

 poured into the flask. The whole apparatus was then steril- 

 ized and used after cooling. Connection was made with the 

 aspirator by means of a strong rubber tube. By referring to 

 the table below, it is seen that the results obtained with the 

 flask method are unsatisfactory, for either no bacteria entered 

 or the number was so large as to lead to a suspicion of con- 

 tamination. 



The flask method was, therefore, abandoned for the filter 

 method, which may be regarded as similar to sugar filters 

 used by Miquel, and also by Sedgewick and Tucker. As 

 sugar is difficult to sterilize, and is also liable to adhere to the 



21 



