OX THE BEHAVIOR OF TUBICOD >l- ANNELID-. 89 



of their occupants have lived and thrived here for more than 

 t \\ enty-five years bears strong evidence to the measure of success 

 in the effort to render them natural. In the-e aquuriu $p:ro- 

 f>his seems to find a fairly congenial environment, and thri\c- 

 continuously in health for many months. For the sake of 

 exhibition advantages the specimens have been planted, or di-- 

 po-ed in such ways as afford the display of the gorgeon-. tl>\\rr- 

 like yill- to the best advantage. Hence some are located on the 

 lloor of the aquaria, others on the back and end- where rocky 

 led^e- afford suitable bases for their support. It oujju aUo to 

 be -aid that in order to render these aquaria the be-i po--il>le 

 exhibition cages the illumination is chiefly, and in -ome case 

 uh'illv, Irom above; while the room itself is purposely kept dark. 

 ex< epi |..r the light which diffuses outward from the aquaria. It 

 becomes important that in reference to the problem before u- 

 thi >f the source and direction of light be borne in mind. 



< in tin assumption of the compelling potency of light it \\ill be 



< le.n that in the case under examination there should be a fairly 

 unilormlv vertical aspect of the various specimens, \\hate\er 

 m. iv ha\e been their original position. The following are the 

 l.it ' ! nn several attempts it was determined with approxi- 

 in, i; i racy that at this time there were about 150 specimen- 



<//>// is in the aquarium. These were di-po>ed, as men- 

 tioned above, on the bottom, ends and back of the tank. < >! the 

 entire number about <)() were in more or less vertical at lit ml 

 \\ith upuard inclination, while 60 were otherwise inclined, tli.n i-, 

 the\ were horizontal or inclining downward. The ^n. ral t 



well shown in Fig. ,v which is a photograph ol the 

 a(|iiarium made by Dr. Sobotta, by whose kind penni--ioii I am 

 able to use it in this connection. Of the 60 specimen- of ihU 

 ad\ erse aspect slightly more than half were horizontally di-po-rd, 

 \\liile the others, some 23 specimens, exhibited de< idedl\ <lo\\n- 

 \\ard inclination. The picture will afford excellent illn-t ration, 

 though not taken at the time my observations \\erc made. 



I el n- now attempt to analyze these facts and their b.-arii 



upon our problem. It may be stated at the out-tart that gra\ - ity 



ha- little or no place in the behavior. I.oeb ha- so concluded 



11 hi- experiments, and my own -o to confirm his venl 



