66 CARNEGIE INSTITUTION OF WASHINGTON. 



Dr. J. W. E. Glattfeld, of the University of Chicago, spent three 

 months as a Research Associate at the Desert Laboratory investigating 

 the most favorable conditions for the reduction of these sugars. He 

 succeeded in preparing a large quantity of xylose of high purity from 

 plant material which has no agricultural value. An extensive search 

 for the best reducing conditions was then undertaken. Various 

 methods were studied for the reduction of xylose to xyUt, including 

 the use of hydrogen with palladium and platinum under various condi- 

 tions, the use of zinc, magnesium, iron, aluminium, and tin under mani- 

 fold conditions of acidity, alkalinity, temperature, pressure, etc. In 

 brief, however, the best results were obtained by the use of 2.5 per 

 cent sodium amalgam. Thus, with the aid of adequate mechanical 

 shaking de\'ices and the proper control of the acidity of the reaction 

 mixture, very good yields of xylit were obtained. Thus far the only 

 dehydrating agent used was formic acid previously treated with phos- 

 phorus pentoxide. This reagent yielded a number of interesting pro- 

 ducts, the nature of which is still under investigation. 



Salt Requirements of Seed Plants, hy B. M. Duggar} 



Two extensive series of experiments designed to throw some light 

 on the salt requirements or "mineral nutrition" of seed plants were 

 conducted at the Coastal Laboratory during the sunmier of 1920. In 

 the main these experiments constitute an entirely new aspect of this 

 work and represent a continuation, on a larger scale in this new direc- 

 tion, of investigations made at the Missouri Botanical Garden in 

 1919-20. As a result of extensive work previously reported (Carnegie 

 Inst. Wash. Year Book for 1919, 84-85; Ann. Missouri Bot. Garden, 

 vol. 7, 1-49, Feb. 1920) it has been apparent to the writer that a "best" 

 salt solution for any particular plant may be regarded as more or less 

 analogous to an " optimum " temperatm-e. It may be affected by many 

 variable factors. However, in the constitution of the solution itself, 

 while the factors of osmotic strength, hydrogen-ion concentration, 

 and physiological balance are all important, they are at the same time 

 variable. This variability is due to the interchange of ions and com- 

 pounds between the roots and the solution vvith lapse of time. In a 

 nutrient solution in which the entire quantity of all solutes is in solution, 

 the variability in the factors mentioned is greatest, and as a rule short- 

 time intervals or frequent renewals of such solutions, involving a very 

 considerable amount of labor, are advised in order to secure the most 

 favorable results. It has seemed desirable, therefore, to ascertain to 

 what extent it might be possible or desirable to substitute for certain of 

 the soluble salts some compounds which are "insoluble" or relatively 

 slightly soluble. Should it be possible to use such compounds in 

 part, or in the main, and at the same time secure the necessary concen- 



* Research Associate of the Department for three months. 



