140 JOURNAL OF THE 



It was the 3cl of December before any further change was 

 observed. At the lower .end the tube was then seen to be more 

 nearly completed on the front side next the glass, and so much 

 of the remaining wall as was visible was closely coated with 

 earth. On the 6th the tube had been lengthened at the upper 

 end, its growth up to this time having been at the lower end 

 only. The addition was about 3 cm. in length and was sup- 

 ported by threads attached to the jar and the piece of weed 

 (Fig. 1). That night at 8 o'clock I went cautiously into the 

 room. Moonlight through a window fell on the jar, and one or 

 two legs of the spider were seen above the opaque part of the 

 tube. On my getting close for better observation, she disap- 

 peared below and refused to show herself again during the half 

 hour that' I waited. She was probably disturbed at her spin- 

 ning, for the tube was open above and, though there was little 

 or no addition to it since I last noticed it, the disposition of the 

 uppermost threads was different. A lighted match held to the 

 side of the jar enabled me on looking down the tube to see the 

 spider below. The walls, slightly sinuous, were quite smooth 

 on the inside. My conclusion that it was open preparatory to 

 further extension was in a measure supported by the fact that on 

 the following morning there was an upward extension of the 

 tube of 5 mm., the end being loosely closed. A little water was 

 poured into the jar to moisten the earth. This kindness seems 

 to have been \vell received, for on the morning of the 9th I 

 found that the tube had lengthened considerably at both ends. 

 The upward extension was 5 cm., turning to the left ; the down- 

 ward, turning gently to the right, was about 5 cm. to the bottom 

 of the jar and thence it followed the wall of the jar along the 

 bottom a centimeter or two, the excavation without silk con- 

 tinuing 3 cm. further. The earth removed had been discharged 

 through a rent near the uj)per end rather on the right side and 

 fell down the slope of the previous dumpings to the left. There 

 was little more excavation after this. Fig. 2 is a sketch of the 

 tube at this stage. 



