132 Active Immunizaüon to Hay Fever [March, 



with ether in water, would give the greatest concentration of pollen 

 grains in the sediment. It was found that a 20 percent Solution of 

 alcohol in water sedimented most pollen, and that, from powdered 

 flowers containing about 8 percent of pollen, a sediment containing 

 15 percent of pollen could be obtained in this way. This method is 

 not satisfactory, however, for the reason that immersion in such 

 Solutions of alcohol, for the time required for Sedimentation, rup- 

 tured the pollen grains, with consequent loss of their contents. 

 Many pollen grains were found, by microscopic examination, to be 

 in this condition. 



The greatest concentration of pollen derived, by the suction 

 method, was about 80 percent for the ragweed powder, in only 

 one sample of 1.75 gm. All other samples contained 50 percent 

 or less. 



Extracts of the pollen were made as f ollows : It was thoroughly 

 triturated with sand in a mortar and treated with a moderate ex- 

 cess of 5 percent sodium chlorid Solution containing 0.5 percent of 

 phenol to prevent putrefaction. This mixture was kept in an incu- 

 bator for 72 hr. at 37° C. and then filtered. None of the extracts 

 by this method gave the biuret reaction and few gave a positive 

 ninhydrin reaction. The filtered extract was then precipitated with 

 8 parts of alcohol and filtered quickly in a Buchner funnel to avoid 

 denaturation, if possible, of the active principle by the strong al- 

 cohol. The precipitate was promptly dried and weighed. This 

 precipitate failed to give a biuret or ninhydrin reaction. It was 

 partly soluble in 0.85 percent sodium chloride Solution and physi- 

 ologically active in very weak Solutions. 



A total content of nitrogen in one of the extracts of ragweed 

 was 0.066 percent. This same Solution, on December 20th, 191 3, 

 gave a positive ninhydrin reaction, whereas on March 24th, 1914, 

 three months later, the test was doubtful. 



The dry precipitate was dissolved in 0.85 percent sodium chlorid 

 Solution with 0.25 percent of phenol, and serial dilutions were made. 

 With these Solutions patients were treated by hypodermic injections. 



The method described above for the Separation of pollen grains 

 from the flowers was cumbersome and the positive results hardly 

 justified the time expended. But the negative outcome in this re- 



