70 Methods in Plant Histology 



not. As a matter of fact, some of the most valuable combinations, 

 like Haidenhain's iron-alum haematoxylin and Flemming's safranin, 

 gentian-violet, orange, require patient study and long practice before 

 they yield the magnificent preparations of the trained cytologist. 

 The beginner, especially if somewhat unacquainted with the details 

 of plant structure, may believe that he has an excellent preparation 

 when it is really a bad, or at most an indifferent, one. To illustrate, 

 let us suppose that sections of the pollen grain of a lily have been 

 stained in safranin and gentian-violet. If the preparation merely 

 shows a couple of dense nuclei and a mass of uniform cell contents 

 surrounded by a heavy wall, the mount is poor. If the two nuclei 

 are quite different and starch grains are well differentiated in the 

 tube cells and the wall shows a violet inline contrasting sharply 

 with a red exine, the mount is good. Anything intermediate is 

 indifferent. If mitotic figures have been stained with cyanin and 

 erythrosin, a first-class preparation should show blue chromosomes 

 and red spindles; if stained with safranin and gentian-violet, the 

 chromosomes should be red and the spindles violet. 



In staining growing points, apical cells, young embryos, anther- 

 idia, archegonia, and many such things, the cell walls are the principal 

 things to be differentiated, if the preparations are for morphological 

 studv. As a rule, it is better in such cases not to use double stain- 



u 



ing, but to select a stain which stains the cell walls deeply without 

 obscuring them by staining starch, chlorophyll, and other cell con- 

 tents. For example, try the growing point of Equisetum. The 

 protoplasm of such growing points is very dense. If Delafield's 

 haematoxylin and erythrosin be used, the haematoxylin will stain 

 the walls and nuclei, and will slightly affect the other cell contents, 

 but the erythrosin will give the cytoplasm such a dense stain that 

 the cell walls will be seriously obscured. It would be better to use 

 haematoxylin alone. For counting chromosomes, it is better to stain 

 in iron-alum haematoxylin alone, or in safranin alone. The same 

 suggestion may well be observed in tracing the development of 

 antheridia, archegonia, embryos, and similar structures. 



In using combinations, it must be remembered that the second 

 stain often affects the first, e.g., if safranin is to be followed by 



