BoT.— Vol. II.] OSTERHOUT— AGAVE. 259 



After being washed from two to eight hours in running 

 water it was transferred to a dehydrator (for description of 

 this see Lawson, 1898, Williams, 1899. or Osterhout, 1900). 

 Dehydration in no case produced shrinkage when allowed 

 to proceed with sufficient slowness. The material remained 

 in the dehydrator until the strength of the alcohol in the 

 upper part was equal to that in the lower. The relative 

 strength of the alcohols may be ascertained as follows : Dip 

 a pipette into the alcohol in the upper part and allow it to 

 take up a few drops by capillary attraction; now dip the 

 pipette into the other alcohol, if on diffusing out of the 

 pipette it falls, it is weaker (on holding it up to the light the 

 course of the diffusion current is easily seen). The mate- 

 rial was transferred from the dehydrator to a mixture con- 

 sisting of equal parts of 95 per cent, alcohol and alcohol 

 from the dehydrator. After standing two hours a part of 

 this alcohol was removed, mixed with an equal quantity of 

 95 per cent, alcohol, and the material transferred to this 

 mixture. It then passed in succession through 95 per cent, 

 alcohol (2-6 hours), absolute alcohol (2-6 hours), absolute 

 alcohol and bergamot oil, equal parts (6-12 hours), and 

 bergamot oil (3-6 hours). It was then placed on the top 

 of the paraffin oven and allowed to warm up slowly, and 

 was then transferred to bergamot oil and paraffin — equal 

 parts — and allowed to remain on the top of the bath. 

 After six to twelve hours it was placed within the oven, 

 and after an hour transferred to paraffin 43° (12 hours), 

 and finally to paraffin 52°, 60° or 72°. 



Sections 1-5 yt* thick were cut on the Minot wheel micro- 

 tome, and fastened to the slide by the water-albumen 

 method. For preliminary examination of ribbons Eisen's 

 alcohol method (Eisen, 1897) proved exceedingly useful; 

 the drying on the oven was omitted, thus effecting a great 

 saving of time without any corresponding disadvantage (the 

 sections were examined in xylene only, in which they do 

 not tend to float off as they do when examined in water, 

 without having been carefully dried on the oven). In view 

 of the fact that the kinoplasmic fibres, and especially the 



