360 BOTANICAL GAZETTE [November 



location or outline of the bundle was wanted, the cross-sections were 

 cut 20 to 30 /x thick; in other cases they were cut 10 ^ thick. Longi- 

 tudinal sections were cut at right angles to the inner surfaces of the 

 cotyledons, as well as parallel to that plane; these were cut 10 \i thick. 

 Staining was done on the slide with safranin and Delafield's hema- 

 toxylin, or safranin and anilin blue. Other stains were tried, but 

 these two combinations gave by far the best results. A few speci- 

 mens were killed by the general picro-mercuric-chlorid method, but 

 no advantage was gained. 



Where seedlings were required, the hard coats of the seeds were 

 cut partly open at the micropylar end, and the seeds placed on one 

 side, partially pressed into moist earth, in a flower pot, and put in a 

 warm moist place in the greenhouse and kept well watered. A good 

 many seeds had been planted three years before and had grown to 

 considerable size, some of them having a stem ioo mm in length and 

 25 mm in diameter, and bearing several leaves. These were treated 

 about as the embryos, except that all the specimens were cut into 

 lengths of 15 to 20 mm . Of course the greater the size of the plant 

 the more slowly the processes of killing, washing, embedding, etc., 

 were conducted. Especial care had to be taken in imbedding the 

 larger specimens, the best results being obtained when they were 

 carried through the process of penetration with paraffin for a whole 

 month. After this time they could be cut with an ordinary Minot 

 rotary microtome with perfect ease, and the sections could be held in 

 complete ribbons. After this precaution, ribbons were obtained from 

 specimens 26 mm in diameter and cut 10 ^ thick. Staining was done 

 as before with safranin and Delafield's hematoxylin, or safranin and 

 anilin blue, the former combination giving the most satisfactory 

 results. An enormous amount of labor is necessary and also 

 much care to keep sections of entire plants in serial order, both in 

 cross and longitudinal sectioning, but it is the only satisfactory 

 method. 



In order to obtain the details or to follow up the leaf trace courses, 

 every section was examined, from the first to the last, and camera 

 drawings made of every second or third section. These were carefully 

 numbered and kept in series and later compared, and reconstructions 

 were based upon them. 



