12 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I26 



Fortunately much more is involved than the mere presence of the 

 mosquito host, as Veloso has demonstrated in his detailed study noted 

 above. To be effective in the transmission of malaria, the bromeliad 

 species must also occur in great quantity within mosquito range of a 

 large human population. Although there is no survey like that of 

 Veloso to give us an idea of bromeliad concentrations in the other 

 states of Brazil, there is reason to hope that the situation in Santa 

 Catarina is exceptional. Certainly nowhere else have I seen or heard 

 of such dense masses of tank bromeliads as occur there. If there 

 remain no serious foci of bromeliad malaria, as seems to be the case, 

 then public health is concerned only with guarding against their de- 

 velopment. The Servigo Nacional de Malaria is already doing this 

 in the south, where I have seen their crews protecting a new beach 

 resort by clearing bromeliads from a belt around it. In the north even 

 this seems unnecessary. Probably the only danger there is the remote 

 possibility of artificially stimulating the growth of bromeliads by pro- 

 viding a favorable habitat, as happened in the development of the 

 cacao plantations in Trinidad. 



Finally, the needs of public health, though paramount, are not 

 wholly irreconcilable with those of horticulture. Destruction of bro- 

 meliads when necessary involves only a narrow belt around a settle- 

 ment and this is not the only means of control nor even the most 

 feasible one in some instances. The species involved are both common 

 and wide-ranging and their extinction is virtually impossible as long 

 as any forest remains. 



PRELIMINARY RECORDS 



All necessary preliminary records have been brought together here 

 and placed in alphabetical order so that they can be more easily con- 

 sulted in monographic studies and so they will not encumber the text 

 when it is used for purposes of identification. 



Aechmea bicolor L. B. Smith, sp. iiov. Figure 100 



A Ae. Candida E. Morren, cui affinis, laminis foliorum valde acutis 

 longioribus angustioribusque subtus omnino cinereo-lepidotis, spinula 

 sepalorum brevi, placentis apicalibus differt. 



Stoloniferous ; leaves in a slenderly crateriform rosette, 5-6 dm. 

 long, much exceeding the inflorescence, covered beneath with fine 

 appressed cinereous scales, sheaths broadly elliptic, 13 cm. long, blades 

 linear, acute with a thick pungent apical cusp, 25 mm. wide, flat, laxly 

 serrulate with teeth 0.5 mm. long, densely cinereous-lepidote above 



