THK IIU-^TS OF N(J\A SCOllA. — FKASEii. 321 



tlie western United States and ^Mexico. The amphispore is 

 evidently a uredospore wbieli has developed the function of s 

 teliosjwre. 



Teliospores. 



These are the last spores of the life cycle. It may possibly 

 be that in some cases they are absent, but usually it is simply 

 a case of not having been found, Teliospores are very varied 

 in their forms and on this account have been used largely for 

 purposes of classification. More stress is now laid on other 

 characters. 



They arise like the urediniospores in sori called telia or 

 teleutosori. Ijeneath the cuticle or the epidermis. They usually 

 break through the epidermis at length, though they may remain 

 covered. The sori may be pulverulent or compact and oft^n 

 dark in colour, though sometimes colourless or bright coloured. 

 They may be borne on pedicels or sessile. In one genus 

 (EndophyUion) not represented in ISTorth America the telio- 

 spores originate in chains and are surrounded by a peridium 

 regarded as the t^st of the telios]>ore. The walls of the telio- 

 spores are usually much thickened and may Ije smooth or 

 verrucose, but never echinulate. Sometimes finger-like pro- 

 jections are i)resent at the apex, as in Piicclnia Lulii. The 

 number of cells varies from one in Uromyces to several as in 

 Phragmidium. The number of germ pores in a cell i^ usually 

 one. In Uroniyces it is always placed at the apex. In the 

 genus Phrar/midiu))} and Gi/ninospDirinr/ium there are several 

 in each cell. 



Th(^ tclia as in P. " )iil)i(io-rera" is often divided into com- 

 partments by modified hyphae, which have been called para- 

 physes, but Arthur regards this as a stroma and not part of the 

 sorus; he regards such sori as compound. 



Some teliospores genninate at maturity <iii the host ])lant. 

 But the teliospore is primarily a wint<;r spore, whose purpose 

 is to tide til" plant over Avintor. and tho majority Avill not 



