KOOT-SYSTKM OF KKAMKKIA CANKSCKNS. 7 



main rang€ <>f the Tucson Mountains. The soil is fine clay mixed with 

 coarse sand, having a depth exceeding 2 meters. The flat is probably never 

 flooded, although for short periods the wash carries much water. In the 

 vicinity are to be found Acacia constricta, Covillea tridentata, F.phcdra anti- 

 svfifii/itica, Parkinsonia torrcyana, Prosopis velutina, /Jzyphus parry i, and 

 other shrubs and trees. 



The specimen of Kramcria whose root-system is to be briefly described 

 was 60 cm. in height and much branched. The root-system of the plant 

 consisted of a very short tap-root and about 6 laterals, which arose from 

 the tap-root near the surface of the ground. They branched very little, 

 and extended in a fairly horizontal direction as far as 2 meters from the 

 central axis. The largest lateral was not more than 5 mm. in diameter, 

 and as it ran outwards was found at depths of 10, 15, 16, and 18 cm. No 

 lateral reached a depth greater than 18 cm. Besides the main laterals 

 there were about as many more slender ones, less than 2 mm. in diameter, 

 which were also placed close to the surface of the ground, but did not 

 reach farther from the main root than 20 cm. The tap-root ended abruptly 

 about 20 cm. beneath the surface. An inspection of the roots showed them 

 to be smooth and lacking the groups of filamentous adventitious roots such 

 as are found abundantly in Franseria, Encelia, and other desert woody 

 plants, which are formed at the beginning of the wet season, generally 

 in summer, and which die with the coming of the dry conditions. 



Several specimens of Kramcria were examined for the purpose of finding 

 a deeply going tap-root, but without success. All of the plants appeared 

 at some time to have formed tap-roots which penetrated the ground far- 

 ther than any seen to be living, but the lower portions of the roots without 

 exception had died. This fact may be connected in some manner with 

 the parasitic habit of the plant, as this precise condition was not seen in 

 any wholly independent plant. Where the tap-root of such forms, i. c, 

 autotropic plants, had died, its place was taken by some lateral which 

 penetrated to an unusual depth. 



It chanced that the roots of the plant first examined did not invade the 

 territory already occupied by the roots of any other plant, and only two 

 roots of other shrubs were seen near the roots of Kramcria, but, in study- 

 ing the root-system of other specimens of Kramcria for the purpose of 

 learning more of the habit of the tap-root, a plant was found of which the 

 roots coincided very closely with the distribution of the roots of Covillea, 

 near which it was growing. In tracing one of the superficial roots of 

 Kramcria it was seen to terminate in a root of Covillea, and examination 

 revealed the presence of haustoria and the fact of parasitism. At the 

 point where the attachment was discovered the root of Covillea was 9 mm. 

 in diameter and that of Kramcria, 1 cm. above the place of attachment, 

 was 2.5 cm. in diameter. On the opposite side and 2.8 cm. farther along 

 the Covillea root another union with Kramcria was seen. Subsequently. 



