55 



Iv 



day. Hydrocotyle umbellata and H, Asiatica were picntiful 

 in the wet sand. The genus Enpatorium was well rep- 

 resented ; for in addition to the species collected the pre- 

 vious day at Norfolk we found E. rotundi folium, E, hyssopi- 

 folium, E. ccelestinum, with its showy blue corymbs, and E. 



r 



fceniculoide s , which was one of the most conspicuous objects 

 in the landscape in places. Whole fields seemed to be in its 

 possession, and the long feathery panicles were singularly 

 graceful in appearance. Trailing over the ground and 

 bushes, and climbing high in trees, we were surprised to 

 find Gelse7)ii}ini sempcrinrens in full flower and with many 

 buds partly opened, although the time of flowering is said to 

 be March and April. Callicarpa Americana^ with its showy 

 red-purple clusters of berries, was a beautiful object, rather 

 sparingly represented. Lechea marituna was collected on 

 the dry sand dunes, and CEnothera htiniifitsa on the sea beach 

 r" -. just above high- water mark. In addition to the above 



J?- 



mentioned, about forty others were obtained, includingr 



<^ 



T 



i 



Rynchospora ghrmeraia, var. paiticidata 



On the 27th we took the railroad to Suffolk, and from 

 thence secured two canoes and boatmen to paddle us up the 

 old canal into the Dismal Swamp. This canal was cut 

 through the swamp for the purpose of floating out timber. 

 It averages some 4 or 5 ft. deep, and about 10 ft. wide, and a 

 journey of about six miles by this means enabled us to form 

 an excellent idea of the vegetation. Throughout several 

 miles there was a dense growth of '^cane-brake'* on each 

 side. {^A ntndinaria jnacrosperma, var. siiffriiticosa.^ The 

 genus Ilex was represented by /. Cassine, L lisvigata and 

 > /. glaber, Smilax laurifolia was omnipresent. Less com- 



mon was S. Walteri, specimens of which were found with 

 light yellow berries tinged with red, for which the name 

 forma pallida is proposed. I was especially interested in 

 noting several trees of Qucrcus heterophylla, not only because 



T 



this increases the range of the tree, but also on account of 

 finding it associated as usual with Q, Phellos and Q. 

 rubra or possibly Q. tinctoria^ between which latter I was 



I 



