502 Hill : Celtis pumila Pursh 



those at Wolf Lake three or four days before, or about April 25th. 

 On May 16th, C. pumila along the Grand Calumet near Miller 

 was visited. It was a little late, the starninate flowers were nearly 

 all gone and the perfect in about the same stage as those of C. occi- 

 dentalis at Maywood on April 29th, showing that their anthesis 

 occurred about the 12th of May. The previous year, May 25, 1898, 

 C. pumila near Miller was examined. The shrubs were well cov- 

 ered by the greenish yellow flowers, the leaves small, from 2-3 cm. 

 long by 1-2 cm. wide. Two days after, May 27th, I was at May- 

 wood and was surprised to see Celtis occidentalis so much more ad- 

 vanced. It did not seem to be due to the mere difference of tempera- 

 ture between the lake shore and the river valley, and offered quite a 

 strong plea for their specific distinction. The fruit on the Maywood 

 trees was well advanced, as large as small peas, or about 5 mm. in 

 diameter, the leaves quite large, 5-8 cm. long by 2—5 cm. wide. 

 On May 18, 1899, C. occidentalis at Wolf Lake was essentially in 

 this stage of development and may be compared with those of C 

 pumila at Miller two days before. The leaves of C. pumila 

 were then 1-3.5 cm. long and 0.5-2 cm. wide. The largest on 

 C occidentalis at Wolf Lake were 9 by 5 cm. These stations, 

 that by the Grand Calumet and Wolf Lake near Whiting, are best 

 for direct comparison. Both are in the sand region of Lake Michi- 



+ 



gan, both subject to the same atmospheric factors of the lake air. 

 The trees at Wolf Lake grow in sand and gravel washed up by 

 waves or heaped by the winds, on rather low land raised 5 to 10 

 feet above the level of the lakelets or bordering marsh, but near 

 enough to water to be moist for trees, being accompanied by elms, 

 basswood, the white and blue ash. The shrubs by the Grand 

 Calumet or on the lake dunes grow in loose dry sand, though 

 their deeper roots get a fair degree of moisture, the surface being 

 from 10 to 50 feet above the river or the level of Lake Michigan. 

 The trees by the Desplaines are in the drift formation, the subsoil 

 a stiff unctuous clay, the surface a clay loam very sticky when 

 wet. The data give a difference of ten to fifteen days between the 

 two species in the sand region and of four to six days in the 

 same species in the two stations, Wolf Lake and Maywood, in 

 1899. It is of interest to note in this connection that one of my 

 time-keepers for the anthesis of C pumila is an orchid, Cjpripcdium 



