The cylindrical cotyledon of A. Dracontmm frees itself from 

 the seed coats and attains a length which varies from 3 mm. to 

 6 to 7 mm. and pushes down into the soil. Before this extension 

 has ceased, the base of the hypocotyl begins to enlarge and in 

 the case of the shorter cotyledons may bring the resulting corm 

 actually in contact with the seed. Coincidently with the swell- 

 ing of the hypocotyl the appearance of two or three roots is to 

 be noted. These are furnished with hairs and are highly con- 

 tractile. In the matter of the development of the plumule the 

 widest variation is shown. In the greater number of instances 

 the plumule is absolutely quiescent during this germination and 

 the formation of the corm goes forward until the seed is ex- 

 hausted, when the roots go into the contractile state and pull it 

 down into the soil with no showing of a leaf during the first 

 stage of its development. This agrees in the main with the 

 behavior of Arum maculatum. In a small number of seedlings 

 of A. Dracontmm, however, the first leaf may be dissected out as 

 a small body about 3 mm. long, of which half is petiole and the 

 other half a rolled green lamina which reaches no greater de- 

 velopment, and never emerges from the cylindrical cotyledon 

 in which it is enclosed. In 9 of the 70 seedlings which 

 came under observation, the first leaf became active before the 

 hypocotyl had doubled its thickness and before more than one 

 root was formed, and extended, forming a petiole 3-4 cm. long 

 and a broad lamina. The seed remains attached to the corm by 

 the cotyledon for an unusually long period and may be seen ad- 

 hering to the corms formed by leafless seedlings in their first 

 resting period. 



The seedlings of Arum maculatum and Arisaema Dracontium are 

 thus seen to be entirely saprophytic during the first season of 

 their development. 



Ignorance of this habit of A. Dracontium led the writer to sac- 

 rifice a fine lot of seedlings of a hybrid between A. Dracontium 

 and A. tripliyllum. During the first season of the development of 

 these plants only seven plumules were counted and when the 

 second season began thirty plants were found, which led to the 

 belief that the culture had been vitiated and the entire lot was 



