| fe So eee eC ee ee ae 
BO =, Saha eee iy 
1913] | HEMENWAY—SIEVE-TUBE 241 
Lupinus polyphyllus, Melilotus alba, Monarda punctata, Oenothera biennis, 
Opuntia Rafinesquii, Phaseolus vulgaris, Phytolacca decandra, Physalis 
heterophylla, Primula sinensis, Sium cicutaefolium, Scrophularia marilandica, 
Silphium laciniatum, Sonchus oleraceus, S. arvensis, Stachys palustris, Tephro- 
sia virginiana, i rinian pratense, Verbascum Thapsus, Veronica scutellata, 
Xanthium spinosu 
The ioe monocotyledons were similarily studied, and all 
were found to have sieve-tubes of the third type. 
Alisma Plantago, Arisaema triphyllum, ieee scandia, Iris versicolor, 
Monstera deliciosa, Polygonatum commutatum, Potamogeton heterophyllus, 
Sagittaria latifolia, Scirpus validus, Smilax eeincnte Typha latifolia, 
Zea Mays. 
There are of course no sharp lines of division in grouping 
these species according to type; even in the same section there 
may be some variation. But in general there is no wide variation 
even with different genera of the same family, except where there 
are both woody and herbaceous genera; then the herbaceous ones 
showed the higher type, as for example in Rosaceae and Legumi- 
nosae. The sieve-tubes of the Leguminosae on the whole are of a 
higher type than those of the Rosaceae. The woody Rosaceae often 
have sieve-tubes about like the first type, with regular large lateral 
sieve-plates. While the woody Leguminosae may have occasionally 
well developed lateral sieve-plates, they are usually as small as in 
Sambucus or Tilia. The xylem of the Leguminosae has been 
likewise found to be a higher type, so perhaps these families do not 
belong so near each other as they are usually placed. 
It will be noted that the woody dicotyledons studied are placed 
in the first and second types, while the herbaceous ones are in the 
intermediate type between the second and third, or in the third 
e. Three or four sieve-plates were the most seen on the end 
wall of any strictly herbaceous plant, as for example Euphorbia 
and Thalictrum. Even in these genera end walls with only one 
sieve-plate were often observed. In the Compositae studied, 
only one sieve-plate to each end wall of the sieve-tube could be . 
found. When stained with Russow’s callus reagent, the pores 
of the “sieve-fields” or lattices showed up as orange dots, but 
they never were large nor fused into callus pads. 
Occasional lateral callus pads were observed in most of the 
