INSTINCT AS TAXONOMIC CIIA K.\( TKK IN SPIDKKS. |,^l 



still more out of proportion with the size of the spider, being 1.2 

 mm. in diameter, with thin, white walls barely covering the 

 dozen eggs. The cocoon was firmly attached to the spinnerets 

 and the spider would not release it when disturbed. On the 

 second day, however, she was found sitting on the bottom of the 

 jar while the cocoon was suspended by a few losely woven 

 threads. On examination, it was found that two of the eggs 

 developed into spiclerlings, while the remainder were parasitized 

 by an insect. Since the spiclerlings emerged from the eggs there 

 is nothing singular in the fact that the mother abandoned the 

 cocoon. 



Here, then, we have proof positive that one of the most stable 

 instincts in spiders may be modified and what is still more in- 

 teresting may be modified in the same sense in two different 

 families, presenting a clear case of adaptive parallelism. It is 

 true that the two families are closely related. In various in- 

 stances species placed in the family Linyphiida? were later trans- 

 ferred to the family Theridiida\ when the presence of the "comb " 

 on the fourth tarsi was ascertained. In the great majority of 

 species the distinction between the two families is quite pro- 

 nounced and certainly there is no close relationship between the 

 genera Bathyphantes and Lithyphantes. 



It is more difficult to decide which of the two methods of 

 taking care of the cocoon should be considered the more primitive 

 one. At first thought it would seem as if carrying the cocoon 

 attached to the spinnerets were simpler than hanging it up in a 

 web. The spinnerets and the chelicerae are the organs employed 

 in the making of the cocoon. The former produce and weave 

 the silk, the latter are used in clipping the sheet and in joining 

 the seam. We know spiders which carry the cocoon in their 

 chelicerae, as for example, Spermophora and Scytodes. The 

 families to which these two genera belong have been often con- 

 sidered to be primitive. It seems to me however, that the family 

 Pholcidae to which Spermophora belongs, shows all signs of 

 specialization, especially in the structure of the cephalothorax 

 and legs. At any rate, we know as yet no truly primitive spiders 

 which would carry their cocoon attached to the spinnerets, and 

 the Lycosickr which generally possess this habit, are not in any 



