286 JAMES CRAWFORD WATT. 



seen which grow partly at the expense of the smaller ones, which 

 redissolve as described for aqueous solutions. But there are many 

 small particles which apparently do not dissolve, but pass on di- 

 rectly over to the larger form of deposit, by forming a nucleus for 

 the laying down of additional material from the clear solution. 



The precipitate exhibits two separate forms. One of these is 

 distinctly crystalline (Figs. 8, 10, 12, 13) and is found most densely 

 at the primary line of contact of the reacting solutions where de- 

 posit was very rapid. Even here the colloid shows its influence, 

 for the crystals are not perfect, but are deformed in various ways, 

 and exhibit rounded angles and suppression of their typical shape. 



The second form of deposit is in the shape of very perfect small 

 spheres (Figs. 7-12, 29, 30, 31), which vary greatly in their trans- 

 parency, markings, and size according to the solution in which they 

 are formed. 



Rainey obtained spherules in gum arabic only after the lapse of 

 an hour or more. At first a faint nebulosity was seen, lasting in 

 thin solutions of gum for one hour, in thick solutions for over a 

 week, after which spherules occurred. In my work I have been 

 able to identify these bodies in less than ten minutes, and in some 

 solutions to see them attain a size of 30 /x in one hour. They grow 

 rapidly and attain practically their full size in forty-eight hours, 

 ranging then according to the solution in which they lie, from a 

 size of 1 /x up to 120^. In Rainey 's case they grew gradually for 

 several months and ranged in size from 2 /a to 200^,. I had one 

 single example of a pear-shaped body which measured 120 jx x 300 fi. 



Evidently gum arabic exhibits a more powerful colloidal influ- 

 ence on the precipitate, but gelatin or albumen will give excellent 

 results in a very much shorter time, which is a valuable consid- 

 eration. 



Brownian movement in parts of some slides is still active after 

 the lapse of a year, due to the colloid retaining some of the pre- 

 cipitate in its original, extremely small, granular form, preventing 

 coalescence to form the larger particles seen in other areas where 

 precipitation was denser and more rapid. 



The process of coalescence of the small particles can be observed 

 in favorable cases. Very small spherules showing active move- 





