Tertiary Spiders and Opilioiiids. 215 



the most misleading methods imaginable. In spiders this is par- 

 ticularly true. We have parallel lines of development in Thera- 

 phosid and true spiders and among the latter in cribellate and 

 ecribellate spiders. The similarity is so great that for years 

 Uloborus and Hyptiotes of the Cribellates were placed among the 

 orbweavers of the Ecribellates. The Agelenids have their coun- 

 terpart in Diplurids among the Theraphosids. The similarity 

 between our common Agelena nacvia and Evagrus mexicanus of 

 the Diplurinse is so striking that it extends over their behaviour 

 and webconstruction, although Ez'agrus is more than twice the 

 size of Agelena and on closer examination possesses all the dis- 

 tinctive characters of the group to which it belongs. Cases are 

 known where external similarity has woefully misled the sys- 

 tematist. Thus a larva of a recent fly belonging to the genus 

 Microdon has originally been described as a hard moUusk, and the 

 cases constructed by some recent caddis-fly larvae as shells. Mis- 

 led by external similarity, the distinguished arachnologist Count 

 Keyserling described a new genus of a cribellated spider. Acan- 

 tlwetentis with two species as a new genus of the sub-order Citi- 

 gradfe, because he did not pay attention either to the spinnerets or 

 the metatarsus of the fourth leg. The synonym}- contained in 

 arachnological works demonstrates clearly that similar mistakes 

 have been made repeatedly by most eminent arachnologists. It only 

 shows that errors are very easy to make and that in the absence 

 of most important characters such errors are inevitable. Under 

 the circumstances a modern arachnologist may well condemn any 

 attempt at classification of extinct spiders. In recent spiders it 

 is a comparatively easy matter to place a specimen in its proper 

 family, it is more difficult to assign it to a genus, and it requires 

 often minute study of characters invisible to the naked eye to 

 determine the species. In this respect the study of extinct forms 

 practically reverses the difficulty. With few specimens in exist- 

 ence it is a comparatively simple matter to establish the fact that 

 a specimen under observation belongs to a dift'erent species from 

 those already known ; but with very few exceptions it is absolutely 

 impossible to place it correctly in the system. What then should 

 be done ? Obviously the species should be carefully described and 

 provisionally placed in such a genus and family with the recent 

 representatives of which it has most resemblance. 



