300 DIFFERENTIATION AND SPECIFICITT OF STARCHES. 



adopting one or the other in accordance with the quantity and physical cliaracters of the 

 specimen. Suflicient water was added and tlie mixture tlioroughly stirred, then strained 

 tlu-ougli several thicknesses of cheesecloth, and then centrifugalized. The supernatant 

 fluid was then removed and fresh water added in its place, and the starch and other matters 

 that had settled to the bottom thoroughly mixed with the water and again centrifugalized. 

 This operation was repeated; or followed by repeated decantation when there was a tend- 

 enc3^ for nuich foreign matter to be tlirown down with tlie starch during centrifugahzation. 

 By this simple process we were almost invariably able to prepare starches that were 

 {)ractically free from gross imjiurities, that is, as free as the nature of our investigations 

 demanded. To have attempted further pmification to the extent of practical demineral- 

 ization would doubtless have proven of disadvantage without proportionate gain. 



PHOTOMICROGRAPHIC RECORDS. 



Verbal descriptions of the histological characteristics of starch-grains fail to con\-ey 

 adequate conceptions, and the same applies to the differences shown by different starches 

 in their polariscopic properties. The notes included in the text have therefore been accom- 

 panied by photomicrographs, both of the grains lightly colored with iodine as seen in the 

 microscope and also in water as seen in the polarizing microscope with crossed Nicol 

 prisms. Absolute reliance can not be placed upon the photomicrographs in showing the 

 differences in the degree of anisotropy of different starches, because the variations in the 

 intensity of the light and in the characters of the negatives or prints may intensify or 

 lessen the effect. The photographs showing the polarizing effects must therefore be checked 

 by comparison with the descriptions in the text, and their usefuhiess is to be found chiefly 

 in showing in each picture the characters of the interference figure, or cross, and chfl'erences 

 in the intensity of polarization of tUfferent grains and different aspects of the same grains 

 of that particular specimen. In making these photographs we used an ordinaiy Bausch 

 and Lomb microscope with a one-fourth objective and a 2-inch eye-piece, which ga\e us 

 a magnification on the field of projection of 300 diameters. Several photographic records 

 are included which show the effects of reagents. 



Details in reference to the values and other features of the methods employed in 

 tills research, and in connection with matters associated therewith, will be found in 

 Chapter VII and in the Prefatory Notes in Part II. 



CURVES OF THE REACTIOX-LNTEXSITIES OF DIFFERENT STARCHES. 



It is excee4ingly difficult to associate the different reaction-intensities of any given 

 starch in such a way as to get a mental picture of their values, and such an association 

 becomes much more difficult if one attempts to duplicate pictures in comparing two or 

 more starches. Hence, the reaction-intensities in relation to each starch and to each genus 

 have been set forth graphically in the forms of curves wliich not only give at once a strik- 

 ingly clear presentation of the quantitative reaction peculiarities, but also permit of the 

 readiest and most satisfactory comparisons. In the construction of the charts (see charts in 

 Part II) the abscissas have been used to express the degree of polarization (P), the intensity 

 of the iodine reaction (I), the intensity of the gentian v-iolet reaction (G V), the intensity 

 of the safranin reaction (S), the temperature of gelatinization (T), the time-reaction of 

 chloral hydrate-iodine (CHI), the time-reaction of chromic acid (CA), the time-reaction 

 of pyrogallic acid (PA), the time-reaction of ferric chloride (FC), and the time-reaction 

 of Purdy's solution (PS). The letter or letters as above given in parentheses each lie at the 

 head of a special cohunn or ordinate, and indicate the agent, while those of the abscissas gi\-e 

 the values of the reactions. The letters of the column under P indicate, respecti\'ely, very 

 high, high, fair, low, and very low; and under I, GV, and S, very dark, dark, fair, light, and 

 very light. The tempei'atures are iu the centigrade scale; and the time-reactions with the 



