Appendix. 



This work was completed at the Anatomical Institute of the 

 University of Minnesota, but the greater part of it was done at the 

 University of Missouri. Some additional observations have been made 

 since the manuscripts were sent to the pubhsher. Better results have 

 been obtained with the simple alcoholic solutions of scarlet red atid 

 Svdan III. It was noted in the body of the paper that these stains 

 give very inconsistent results, but the cause of this variability was 

 not clear at that time. The staining-power of these solutions depends 

 apparently upon the degree of saturation of the alcohol with the dye. 

 A stain may be weak because it was not thoroly saturated when pre- 

 pared; and a strong stain may weaken upon standing because of pre- 

 cipitation of the dye. If a clean vial be filled with a freshly-filtered 

 highly-saturated solution, it will usually be noticed that a precipitate 

 soon collects upon the sides and the bottom of the vial. In proportion 

 to the accumulation of this precipitate the stain becomes weaker. 

 For this reason a stain that has been in the laboratory several days 

 is apt to be considerably weaker than a freshly-prepared solution. 

 The stain must be at its maximum efficiency to color the faintly- 

 refractive liposomes. 



The quality of the dye employed seems also to be a factor of some 

 importance. Only Grûbler's dyes have been used, but with some 

 samples it seems impossible to prepare a simple alcoholic stain strong 

 enough to bring out the faint liposomes. This difficulty was not en- 

 countered in preparing the Herxheimer solution. 



Very efficient solutions have been prepared by using a large 

 excess of the dye (2 g to each 100 cc of 70 or 80 per cent alcohol) and 

 shaking vigorously fifteen or twenty minutes. It is safer to heat the 

 alcohol (as recommended by Fischer), but equally good results have 



