8 University of Michigan 



The distribution of Hetaerinas on a particular stream may 

 be such as to still farther complicate any ideas we may have 

 as to the manner of distribution. For example, suppose there 

 are several small ' streams flowing in the same direction from 

 the same range of hills and crossing a narrow coastal plain 

 to enter the sea. It is obvious that the streams will approach 

 their neighbors on either side very closely at their ramified 

 heads. Moreover, these headwaters will approach very closely 

 the headwaters of streams flowing in the opposite direction 

 from the range of hills. Given a species of Hetaerina on one 

 of these streams, it is obvious that it might reach adjacent 

 streams across the coastal plain or by the narrowly divided 

 headwaters, which would also permit the extension of the 

 species to the opposite side of the range of hills. These two 

 courses offer apparently the easiest paths for the dispersal 

 of the species. 



But the problem is not so simple. Let us look at the Rio 

 San Esteban in Venezuela. Where it leaves the rocks to flow, 

 still swiftly, in its sandy bed across the coastal plain, Hetae- 

 rina caja is abundant. Upstream among the rocks and in the 

 lower courses of the tributary quebradas Hetaerina macropiis 

 is found. And at the heads of these quebradas, where the 

 rock masses are the roughest and most precipitous, lives the 

 largest and handsomest of the San Esteban Hetaerinas, capi- 

 talis. And on the opposite side of the mountains at Bejuma, 

 for example, in the Orinoco drainage, we find the same distri- 

 bution of the same species. Furthermore, at Maraquita, far 

 away in the interior of Colombia, seven or eight hundred 

 miles up the Magdalcna River, we find the same three species 

 with the same distribution. 



We may assume that these three locations, with streams 

 of similar character, have been in the line of similar flows of 



