Lange : REVEGETATION OF TRESTLE ISLAND. 623 



species is very common along the railroad track south of the 

 island. These prothallia were little green clumps from */ to 

 y^ of an inch in diameter and grew mostly on the walls of 

 the shaded mud cracks. From many of them one or more 

 thin young horsetails were protruding. Although I carefully 

 searched the cracks for these prothallia in the springs of 1899 

 and 1900, I did not again find a single specimen. In 1898 

 they were abundant on both sections of the island, but they did 

 not grow on the marl, but only on silt and in silt cracks. This 

 silt consisted of a brownish clay and of very fine quartz sand, 

 making a compact damp soil. 



The West section was still quite bare.early in July, 1898. The 

 creviced marl ridges had dried in the sun and exhibited white 

 streaks like limestone. These little crags and points were bare, 

 but in shaded and sheltered depressions mosses covered much 

 of the damp marl ; and small cottonwoods, aspens and willows 

 were scattered over the whole West section. 



About September loth the aspect of both sections had 

 changed, but most markedly that of the East section, which 

 was a veritable wilderness of weeds. Its lower portion was 

 covered with wild rice, over six feet high. Under the rice and 

 also on the higher ground water hoarhound (Lycopus ameri- 

 canus], skullcap (Scutellaria laterifolia) and American wild 

 mint (Mentha canadensis] formed dense tangles and grew with 

 a luxuriance I had never seen before. There were present in 

 great confusion nearly all the plants mentioned in the count of 

 August 16, 1899, an d a f W not found at that day. In this 

 weeds' paradise the young willows could be seen, but they were 

 not at all conspicuous. No bare ground was now visible on 

 this section, but in walking through the weeds, one could not 

 avoid frequent stumbling into the cracks. The West section at 

 this time presented the appearance of a loose growth of young 

 cottonwoods, aspens and willows, with the cottonwoods most and 

 the willows least conspicuous, but much bare soil could still be 

 seen even from a distance. 



On both sections all the young trees remained green and con- 

 tinued to grow until late in the fall. On October 22d they were 

 still green but had evidently ceased growing on account of the 

 cool weather that had prevailed since the third of the month. 

 Of the East section my notes for October 22d say: "A great 

 wilderness of dead weeds." 



