12 BULLETIN 1057, U. S. DEPARTMENT OF AGRICULTURE. 



as the tree grows along the banks of this klong, the waters from 

 which must inundate the surrounding land in the rainy season. It 

 grew in company with Dipterocarpus alatus, Melia arborea, Salix sp., 

 and other trees of which specimens have been collected but are as yet 

 unidentified. The leaves of wild-growing Hydnocarpus anthel- 

 minthica are not as narrow as those of cultivated specimens, but are 

 lanceolate instead of linear-lanceolate (PI. II). No mature fruits 

 were found, but semiripe fruiting specimens were collected. 



The natives of Siam use the seeds of this species for the expression 

 of an oil which they apply in cutaneous affections. It was stated 

 by natives that the fruits are poisonous and also that fish feeding 

 on the seeds of maikrabao are unpalatable. The narcotic properties 

 of the seed are imparted to the fish, which, when eaten, produce 

 nausea and vomiting. 



The seeds form an article of export to China, where they are 

 known as " ta-f ung-chi." While in Hongkong the writer inquired 

 about ta-fung-chi and found them on sale in native Chinese drug 

 stores at $20 for 100 pounds. Since only about 32 per cent of the 

 seed is kernel, the price is rather high. 



HYDNOCARPUS CASTANEA." 



While in search of Taraktogenos hurzii near the Siamese border 

 but in the Burmese territory, the writer came across a species belong- 

 ing either to Hydnocarpus or to Taraktogenos. It grew in a dense 

 rain forest, on steep, densely wooded slopes, near the jungle village 

 of Thinganyinon. Unfortunately, the tree was not in flower, but it 

 had immature fruits, a comparison of which with those of Tarak- 

 togenos Jcurzii rendered doubtful its identity with that species. The 

 time of flowering is in April, and the fruits ripen in July. The 

 writer's visit was in December. 



Having known that Taraktogenos hurzii was said to occur in the 

 Martaban Hills, the writer inquired, on reaching Moulmein on De- 

 cember 24, 1920, whether anyone had seen kalaw trees anywhere in 

 the Martaban Hills. A Mr. Shwaloo, the son of a Burmese physi- 

 cian, stated that he knew where there were thousands of kalaw trees. 

 Arrangements were made to explore the Martaban Hills, and a 

 party, consisting of the writer, Mr. Shwaloo, a cook, and a boy, 

 left for Martaban station and thence by rail for Paung. From 

 Paung the party proceeded by bullock cart to Oktada, a small vil- 

 lage at the foot of the Kalama Range. The same evening the writer 

 followed the steep dry creek bed, strewn and lined with enormous 

 bowlders of quartz rock, to the Mondo Range. In the crevices be- 



• Hydnocarpus castanea Hook. f. and Thorns. (12, p. 197). Specimens collected by the 

 writer are deposited in the United States National Herbarium : Rock No 744 U S N FT 

 No? t.090,008 and 1,090,009. ' ' ' ' ' 



