4I>^ 



NA TURE 



[February 28, 189- 



make it possible to attempt to weave them into a 



coherent and intelligible whole, by tn-ing to trace the 

 origin and evolution of the habit of net-soinning. It is 

 strange that but a small number of students seem to 

 have occupied themselves with this most attractive 

 aspect of the subject. With the exception, indeed, of a 

 few authors who have here and there thrown out stray 

 suggestions upon particular points, no one appears to 

 have seriously set himself to the elucidation of the whole 

 problem. It is true that in the second volume of his 

 work upon the American orb-weaving spiders, Dr. 

 McCook devotes a chapter to the "genesis of snares" ; 

 but since he does not appear to be able to attach great 

 importance to the evidence in favour of evolution, his 

 treatise on the subject practically resolves itself into a 

 demonstration of the fact that, by starting at any point 

 you please, in what is called " aranead spinning work," a 

 series of gradations may be traced from one modifica- 

 tion of architecture to another, from the simplest to the 

 most complex, or from the most complex to the simplest. 

 He thus succeeds in leaving his readers completely in 

 doubt as to whether or not he intends one or all of his 

 attempts at tracing the "genesis' of snares to represent 

 what has actually occurred in the course of nature; and 

 one closes the chapter without satisfactorily ascertaining 

 if its writer has any definite views respecting a primitive 

 form of spinning work. 'Set, at the same time, it must 

 be admitted, an impression remains that the suggestions 

 that are put forward, based as they are upon an extensive 

 knowledge of the subject, point in more than one instance 

 to the true lines along which the web spinning habits 

 have been evolved. 



In attempting to arrive at an understanding of the 

 origin of any structure or instinct in an animal, one 

 nowadays naturally refers for an explanation to what is 

 hypothetically its ancestor, or, failing this, its ancestor's 

 nearest ally. If this method of research be adopted in 

 connection with the spinning powers of spiders, it is 

 found that silken threads are fabricated by two allied 

 groups of animals, both of which are believed by some 

 students to stand, in many respects, nearer than spiders 

 do to the ancestor of the class to which spiders, scorpions, 

 mites, &c., belong. In one of these — the Chelifers, or 

 book-scorpions — the presence of silk glands has long 

 been known. In the other — the /"//rj/i/V/o-- their exist- 

 ence is now, for the first time, I believe, pointed out. 

 Thefunction of the silk in the Chelifer is cocoon-spinning ; 

 and that it is materially the same in the I'hr\nid<r is 

 shown by the easily verified fact, that the egg-case of the 

 mother is secured to the lower surface of her abdomen 

 by fine silk-like threads. One of the chief interests of 

 this discover>' lies in the circumstance, that of existing 

 animals the Phrynidir appear to be most nearly allied to 

 the immediate ancestor of the spiders. We are, there- 

 fore, justified in concluding that originally the silk in 

 spiders was utilised for the purpose of making a case for 

 the eggs. 



If, however, we consider the question from the stand- 

 point of spiders alone, it seems to me that we should 

 naturally arrive at the sa'ne result. For it n, ^ priori, 

 probable that the primitive form of spinning industry was 

 that particular kind which is now common to all groups. 

 But when we pass in review the spinning work of the 

 various tribes of spiders, we find that the habits of 

 titilising the silk for constructing a snare, or drag-lines, 

 for enswathing captured prey, or for purposes of locomo- 

 tion, do not occur, by any ineans, invariably throughout 

 the class. In fact, we cannot say of any one of them 

 that it is characteristic of spiders. Not so, however, is 

 it with cocoon-spinning. For, however dilTcrenl from 

 each other m structure spiders may be, and however dis- 

 similar in habits and mode of life, we yet find that the 

 instinct of the mother to spin a cocoon for the protection 

 •of her eggs is never wanting. 



NO. 1322, VOL. 51] 



Granting, then, the possession of silk-glands inherited 

 from an ancestor, we may conclude that the first step in 

 the development of web-spinning was the formation of 

 the cocoon. What was the second .' We know that a 

 spider's care for her eggs does not, as a rule, cease with 

 the completion of the cocoon : some species carry it 

 about with them ; others mount guard in its vicinity. 

 Possibly the former was the original method of disposing 

 of it. But if so, since such a habit must more or less 

 impair the mother's activity and must render her a con- 

 spicuous object of attack, we can understand why it has 

 been abindoned for the latter method by the great 

 majority of spiders, and is now almost confined to those 

 species in which the nomadic mode of life reaches its 

 highest development. If, on the other hand, as seems 

 more likely, the primitive habit was that of watching by 

 the cocoon, we can understand that during the tem- 

 porary period of quiescence thus enforced, the mother 

 would naturally seek concealment and protection for 

 herself; and since she possessed the instinct and material 

 for constructing a receptacle for her eggs, it is possible 

 to see how a slight modification of intelligence might 

 have led her to extend the same protection to herself by 

 weaving a covering over and around the retreat in 

 which she had sought refuge. Then if an aperture for 

 ingress and egress, for purposes of feeding, were left at 

 any spot in the wall of such a protective domicile, there 

 would arise, in a rudimentary form, what is known as the 

 tubular nest or web. .\nd the next simple but important 

 step would doubtless be the adoption of the silken tube 

 as a permanent abode for the mother after the dispersal 

 of the young to shift for themselves. 



As a matter of fact, some spiders have advanced no 

 further than this stage. The females of some Drnssiiiic, 

 for instance, spin a temporary retreat for themselves and 

 their young at the breeding season ; while others utilise 

 the retreat as a permanent dwelling-place. Lastly, the 

 view that the forination of a tubular retreat was in re dity 

 the second stage in the evolution of web-spinning, seems 

 supported by the circumstance that the tube, whether 

 accompanied or not by accessory developments, is, with 

 the exception of the cocoon, the most constant feature in 

 the spinning industry of spiders. 



Adopting then, for these reasons, the conclusion that a 

 simple tube was the priinitivc form of nest, it seems that 

 the evolution of web-spinning has been carried out along 

 two main lines. Along one there is a gradual elaboration 

 of the tube until it culminates, so far as structural com- 

 plexity is concerned, in the trap-door nest with which 

 everyone is familiar ; along the other, the tubular nest 

 either ultimately disappears, or, retaining its primitive 

 simplicity, it is to a greater or less extent superseded by 

 the formation of a new structure — namely, the net 

 for ensnaring prey. 



It will not here be necessary to enter upon a discussion 

 concerning the various forms of tubular nests that are 

 constructed ; but a few words respecting the probable 

 origin of the door-making habit may prove of interest. 



In the first place, it is important to note that the remark 

 able instinct to close the aperture of a tubul.ir nest with 

 a movable lid is possessed by spiders belonging to two 

 groups. These are \Mc I.ycosuia:, or wolf-spiders, of which 

 the .South European 'I'aranlula is a historical example, 

 and the gigantic Aviciilariidr, which have won such a 

 bad name for their alleged bird-catching propensities. 

 But although there is no direct genetic altinity be 

 tween the species coinposing these two families, it i> 

 nevertheless highly interesting to note that they present 

 a close parallelism in nest architecture. In both there 

 arc species which form no nest, others which construct 

 a simple silken tube, and others which close the 

 aperture of the tube with a hinged-door. Yet it is certain 

 that the last-named instinct has been independt-nili 

 acquired in the two cases. Moreover, it is probable, a!> 



