(> liiilliliii Atiiiiicdii .U(/.s((/w of Xdlitrul Hidorij. [\'ol. XXIX, 



Dkpaimikks fko.m Simon's Wokk. The Introduction of a New 

 Tkkm. (Oiiohs. — Sonu" (U'paitiircs were, liowever, fouiid to he necessary 

 ;iii(l tlu'v should he incntioncd l-',\;miininjf the spiders of the whole world 

 ;iiid liiidinji many iiitcr^rachiig species Simon was forced to throw together 

 some genera into single pro\isional groups to which he gives the position 

 of genera, l)Ut which an- in reality considerably more comprehensive. 

 They may l)e regarded as a first means of avoiding difficulties, conglomer- 

 ates of genera with distinct types and intergrading species. To such 

 conglomerates I propose tlie name Cohors, it being a short word with a 

 meaning somewhat different from that understood by supergenus, the 

 latter term having a definite position in the system. All the old genera of 

 Walckenaer, Koch and others may be at present regarded as such cohortes, 

 but tlie majority of tliem have been gradually split into definite genera. 

 Still se\eral remain. The genus Ericioue as used by Keyserling is in my 

 sense a cohors. Simon di\ided it many years ago into a number of real 

 genera accepted in the present catalogue and limited the genus Erigone to 

 forms possessing only certain characters. But unfortunately the species 

 described by Keyserling cannot be placed in Simon's system from the mere 

 descriptions and I had neither time nor opportunity^ to examine those of 

 his types which still exist. Under the genus Erigone will therefore be found 

 a great number of forms that do not belong to it and to facilitate future 

 work for the student I have in each case mentioned whether or not the 

 genus is correct. In the family Argiopidpe, Simon's Araneus is a cohors. 

 F. Cambridge has made an attempt to split this cohors into several true 

 genera but his work was only tentative and his genera cannot be ranked 

 with such as are well established. Some future student will have this 

 enormous task to accomplish. Meanwhile I thought it wise to place all 

 genera of F. Cambridge under the cohors Araneus, but also to indicate his 

 genus l)y putting it in parentheses. All species referred l)y earlier writers 

 to Walckenaer's genus Epeira and not yet correctly placed, will likewise be 

 found under Araneus. Some of them may prove not to belong to the 

 undergroup Araneinjc of Simon and be placed eventually under Mangora, 

 Ci/closa, etc., as in the case of Mangora (Epeira) gibherosa Hentz. In the 

 family Lycosidse Lycosa forms a cohors, but since this family contains only 

 a few genera, I left the genera Alopecosa, Allocosa, Schizocosa, Arctosa, etc. 

 I do not suppose that these genera will all withstand criticism, perhaps with 

 the single exception of Sekizocosa, but I did not want to add new names 

 where it was easy to avoid it. Finally, in the Salticidse there remain two 

 cohortes, Aftus and Dendryphantes. The species belonging to the former 

 cannot at present be properly classified and had to be left as they were. 

 ^^ ith the cohors Dendryuhantes it was somewhat different and here I fol- 



