VALUES OF METHODS IN THE DIFFERENTIATION OF STARCHES. 311 



GENERAL CONSIDERATIONS. 



The methods used in this research in the differentiation of the starches of different 

 plants and phmt i)arts, and especially in indicating different forms of the starch-substance, 

 are exceedingly crude, and, as a consequence, the range of error of experiment is in 

 comparison with chemical jirocedures generally very large. The differentiation of isomers 

 at the present time depends, as {)rcviously stated, upon methods which, in a very large 

 measure, are crude and quantitative in nature; as, for instance, differences in density, 

 solubility, decomposability, color and color reactions, digestibility, toxicity, physiological 

 pi'opcrtics, temperatures of gelatinization and melting-point, optical reactions, crystalline 

 forms, etc. The closer the relationship of isomers the greater usually the difficulty in their 

 differentiation; hence while little or no difficulty may be experienced in differentiating the 

 isomers constituting a group of polysaccharoses, or disaccharoses, or monosaccharoses, the 

 differentiation of corresponding substances, or stereoisomers, such as different forms of 

 starch, glycogen, hemoglobin, etc., may be attended with great obstacles. 



In the case of many isomerides which are optically active and crystallize in rectilinear 

 forms, as, for instance, the hemoglobins, accurate differential and diagnostic data can be 

 obtained, but in case of starch and starch-like substances the conditions are obviously 

 quite different. Were it not that the starch-grain during its growth is subjected to various 

 mechanical and changeable conditions which affect its form, and that different plant jiarts 

 may not gi\'e rise to the same form of grains because of different forms of plastids or other 

 conditions, and that the grains vary in form at different stages of development, the starch- 

 gi'ains of any given plant would in all probability be of a single histological type and as 

 specific in relation to the plant as the offspring to its parents in other words, its form 

 would be an index of the peculiar biology of the plastid (were there a single form of 

 plastid), and as this plastid differed from the plastid of another plant the product would 

 likewise differ in its microscopical properties ; but having different forms of plastids in the 

 same plant, and variable conditions, the forms of the starch-grains are corresjiondingly 

 variable. 



These differences in conditions give rise not only to the presence of a variety of forms 

 of starch-grains, but also to variations in the proportions of the different forms in any 

 given starch. It therefore follows that if a number of preparations of starches were obtained 

 from as many plants of the same species, no two need be exactly alike, either in the rela- 

 tive proportions of different types of grains or in the characters of the most conspicuous type 

 of grains present. Moreover, owing to this heterogeneous and variable mixture of shapes 

 and sizes, no two fields seen on the stage of the microscope are ever identical, and sometimes 

 they are so different as to suggest that the fields represent starches from different species. 



In comparing the plates with the text such discrepancies may be apjiarent in a seem- 

 ing lack of absolute agreement of text with picture. Therefore in the descriptions of the 

 grains the impressions given are those obtained by a general examination of what might 

 be expressed as the exhibit of the starch as a whole. Similarly, variations in the position, 

 form, and other characters of the hilum, variations in the lamella?, variations in fissura- 

 tion, and variations in the number and character of aggregates and compound grains, etc., 

 are likewise to be expected. Moreover, no two persons would in the least likelihood be 

 apt to give the same description of any given kind of starch, as is instanced in the records 

 in various parts of this work, as in Maranta arundinacea (page 224 and Part II, 813), 

 and also in the warious popular works treating of starch-grains. It is therefore ob\-ious that 

 the gross histological descriptions of starch-grains can not be laid down on hard-and-fast 

 lines as in the case of ordinary geometric forms. 



Again, in the reactions the personal factor plays a very important part on the records. 

 The intensities of the color reactions are based upon an arliitrary standard of the experi- 



