98 Mr. J. Nixon o?i the Measurement q/'the 



of air is usually seen to envelop the abdomen. Misled by the 

 appeai'ance of the bubble, in the case of the house-spider, 

 which was the subject of his experiment, the author of the 

 article in the Zoological Jouinal endeavours to account for 

 this fact on the supposition that the air is contained in a web 

 formed about the body of the spider with the assistance of the 

 hinder legs. Those zoologists who are familiar with the ex- 

 ternal structure of spiders will immediately perceive the in« 

 sufficiency of this explanation: it will be equally apparent also 

 to the uninitiated, when they are informed, that the cephalo- 

 thorax and even the legs individually, as well as the abdomen, 

 are frequently encompassed with air. In short, the air, which 

 in the form of bubbles envelops to a greater or less extent the 

 body and limbs of spiders when immersed in water, — instead 

 of being contained in a web, is confined among the hair with 

 which those animals are clothed ; consequently, however sud- 

 den or unexpected their submersion may be, they are always 

 prepared for it, as I have ascertained by repeated trials. 



With these results before us, it will be readily admitted, 

 that in conducting experiments with spiders placed on objects 

 insulated by water, there is a decided advantage in employing, 

 as I uniformly do, vessels having smooth perpendicular sides; 

 care being taken not to Jill them with the liquid. Moreover, 

 when the experiments are made with hunting spiders, a vessel 

 of considerable internal dimensions should be selected; for if 

 this precaution be neglected, some species, Salticus sceniciis for 

 example, will escape by leaping over the water intended to 

 confine them ; and as on such occasions a line attached by its 

 extremity to the station previously occupied by each indivi- 

 dual, is drawn out after it from the spinners, the notion that 

 it had been wafted to the edge of the vessel by a current of 

 air, might be induced in this case, as it was in that of the 

 house-spider, to an exposition of which so large a portion of 

 the present communication is devoted. 



XXII. Particulars of the Measurement, by xmrious Methods, 

 of the Listrumental Error of the Horizon-Sector described iii 

 Phil. Mag. vol. lix. By John Nixon, Esc[.* 



[Continued from Phil. Mag. and Annals, N.S. vol. x. p. 347.] 



By the Sixth Method. 

 Theory. — HPHE elevation of a mountain, observed from a 

 -^ station at its base, should be found, on trans- 

 porting the instrument to its summit, to be equal to the cor- 



* Communicated by the Author. 



