Bibliographical Notices. 101 



luros ill tlio poiierally-accepted scheme of classification of the order 

 Araiieie. This .scheino, of Avliich Dr. ThorcU is the most able 

 cx])oiiei)t, depends upon the fact that a classification of the 'webs 

 accurdinf^ to tlieir form corresponds closely witli a classification of 

 the spiders based upon the sum of their most obvious structural 

 features. AVith the rival scheme *, which is established upon the 

 existence in otherwise dissimilar genera of those curious orf,'an.s 

 known as the cribdlnm and calcDiiistrum — a scheme for which 

 Dr. Eertkau has said all that is to be said — we need not further 

 deal. It will be sufficient to state that Dr. ilcCook, rightly in our 

 opinion, adojits the views of Dr. Thorell, and associates with the 

 ()r])itelaria) the aberrant genera Uloborus, llyptiotes, and Theridio- 

 soma. 



liut a noticeable circumstance connected with this matter is that 

 although, as above stated, a natural classification of the wel>8 

 coincides with a natural classification of their makers, when the 

 Aranoa; as a whole are considered, yet the principle is found not to 

 apply if an attempt be made to extend it to the suborder now under 

 discussion. In other words, an obvious classification of the snares 

 of the Orbitelariaj does not correspond with a classification of the 

 species and genera according to their affinities as exemplified by 

 structure. As an illustration of this may be pointed out the fact 

 that within the limits of the genus Epeira webs of very different 

 types may be constructed. The commonest type is a simple, vertical, 

 full-orbed net with a meshed hub {$ic) ; but in the species known as 

 Ep. labyr'mthea a system of netted lines is associated with the ordi- 

 nary web ; in Ep. triaranea the web is not full-orbed, but lacks one 

 sector ; the web of Ep. gihherosa is horizontal and not vertical ; 

 and, lastly, Ep. basilica weaves the remarkable net which Dr. 

 McCook has described as the domed-orb. On the other hand, the 

 web of Uastracantlia is almost like the web of the ordinary type of 

 Epeira ; that of ZiJla., not to mention JVqjhila, resembles that of 

 triaranea in lacking a sector : that of Teira<jnatha is like that of 

 gibberosa in being horizontal. It appears, then, that there may be a 

 greater difference between the webs of a species of a genus than 

 between the webs of distinct genera ; thus the web of Epeira basilicct 

 is far more unlike the web of, e. g., Ep. diaderaata, than is the web 

 of ZiVa or even Arr/io^^e. 



Since, then, the form of the web is liable to so much variation 

 within the limits of a single genus, and since species belonging to 

 different genera may spin snares that are almost exactly alike, it is 

 clear that great caution should be used in concluding that spiders 

 ■which make webs on a particular plan are necessarily related to 

 each other. But it is impossible to pursue this interesting topic 

 further. Enough has been said to give some idea of, perhaps, what 

 is one of the most important lessons to be learnt from Dr. McCook's 

 researches into the nature of webs. 



* For an able and exhaustive criticism of this classification refereace 

 mav be made to Dr. Thorell's paper in the Ann. & Mag. Nat. Hist. 

 voh xvii.pp. 301-31^0 (1886). 



