166 PEl'PEK TREE AS A CAUSE OF HAY FEVER. 



The scarcity of pollens other than that of the pepper tree in 

 the air of the town is probably to be attributed to the distribu- 

 tion of the rainfall and the nature of the vegetation in the vicinity 

 of the town. The vegetation to the west of the town (the direc- 

 tion from which the prevailing strong winds blow during the 

 epidemic season) is almost pure grass, the flowering of which is 

 determined by rain. Though growing rains may fall in the early 

 summer, it is usually not until February that the rain is suffi- 

 cient to enable the grass to flower on a large scale. The vegeta- 

 tion of the kopjes to the north and north-east of the town is more 

 varied, and includes some shinabs whose flowering season, though 

 also dependent on the rain, is less so than that of grass ; but 

 strong winds from this direction are rare. 



The weather, apparently chiefly the rainfall, also exerts a 

 great influence on the date and profusion of flowering of the 

 pepper tree, an exotic which is cultivated in gardens and in the 

 streets. In the three seasons 1919-1921 it came into flower to- 

 wards the end of October, and simultaneously epidemic hay- 

 fever commenced. The principal flowering season of the pepper 

 tree is November and December, and it is in these months that 

 the .epidemic is most severe. 



A reference to Table II will show that the hay fever 

 epidemics occur towards the end of a long dry season. Thus, 

 in 1919 the total rainfall at Bloemfontein for the six months 

 May — October was only 1-96 inches. 



Though the results of an examination of the air of Bloemfon- 

 tein for three consecutive epidemic seasons are in general agree- 

 ment in emphasising the presence of dust and pepper tree pollen 

 and the virtual absence of other pollens, it is probable that obser- 

 vations over a series of years would show the kinds and quantity 

 of native pollen in the air at this season to vary from year to 

 year, and to be determined chiefly by the rainfall. Thus, with 

 heavy rains in early spring, extensive areas of grass will flower in 

 December, and grass pollen would then be expected in the air. 



It should be noted that no pollen of any of the American hay- 

 fever w-eeds (Scheppegrell, 1916) was found on the pollen plates. 

 Many species of them occur as weeds in and around the town, 

 but they never flowered freely during any of the three epidemic 

 seasons in which pollen plates were exposed. Thej' are mostly 

 annuals whose flowering, like that of grass, is dependent on the 

 rains. With repeated heavy spring rains they, too, might be ex- 

 pected to flower during the epidemic hay fever season. 



IV. Is Pepper Tree Pollen Capable of Causixg Hay Fever? 



This was tested in two ways : (a) By what is known as tht- 

 Cutaneous Test. This is the standard method of testing suscepti- 

 bility to a particular pollen, and consists of scarifying the skin, 

 of, say, the arm, and applying to it the neat pollen. The reaction, 

 if any, is local. In nature, hay fever pollens are inhaled, but 

 patients hesitate to inhale pollen in an experiment, as the resvilt, 

 if the patient is susceptible to the variety of pollen used, is to 



