SPONGES IN WATERWORKS. 975 



{cf. Kraepelin, loc. cit.). It can grow as well in the tlai-k, with- 

 out zoochlorellte, as in the light ; and being eflfectively protected 

 from frost and other adverse conditions, it seemed possible that 

 it might continue alive during the whole winter. On the same 

 day, we made a cursory examination at Llanishen and Lisvane, 

 with no result ; but on the following day, the reservoir-keeper, 

 after emptying the valve-shaft at Llanishen, found a considerable 

 growth at a depth of between 10 and 20 feet. 



Since then, a careful look-out has been kept for the sponge 

 throughout the entire system, but there has been no positive 

 evidence of its occurrence in other jjarts, either in the open or in 

 the underground pipes. 



The tw^o most important problems which presented themselves 

 were — the destruction of the sponge in the infected area, and the 

 prevention of re-infection tlu'oughout the whole system. As 

 regards the latter, the only sound method would be the intro- 

 duction of pre-filtration through sand, so as to starve the sponge 

 by preventing the mici-oscopic organisms (on which it and other 

 " microphagous " organisms which might at any time appear 

 depend for their nutriment) from passing into the pipes and 

 reservoirs (cf. Kemna, quoted by Harmer on p. 432), as well as 

 to keep out larvte and gemmules. This method would take 

 a long time to carry out and entail considerable expense ; 

 moreover, it would not be effective unless it could be adopted at 

 the lower reservoirs (where there are difficulties as regards level) 

 as well as the higher ones, as there would be a reappearance of 

 organisms in them. As the matter was urgent, I therefore felt 

 it would be better to confine attention at first to the destrviction 

 of the organism in the infected area and the prevention of its 

 regeneration and of reinfection by the gemmules. 



Treatment with any destroying agent except such as would be 

 dangerous to human life would be ineffective if applied to the 

 adult sjoonge, as it would not destroy the myriads of gemmules 

 protected by their resistant coats. It was therefore necessary in 

 the first instance to make a periodical examination of the sponge 

 for a year or more, so as to ascertain (1) whether sexual repro- 

 duction occurred, (2) when the gemmules germinated, and 

 (3) whether the adult growth perished partially or entirely 

 during the course of the winter*. 



An examination of specimens at various times of the year has 

 so far revealed no larvse or any signs of sexual reproduction ; but 

 the enormous production of gemmules throughout the crusts and 

 branches is quite sufficient to start new growths in any part to 

 which the gemmules could gain access. There can be no doubt 

 that a sexual stage is unnecessary for the repi-oduction of the 

 sponge from year to year. 



* Cf. Marsliall (SB. Natuvf. Gesellsch., Leipzig, 1884, & Journ. Roy. Micros. 

 See. V. 1885) ; Weltner (Arcliiv f. Entwicklurigsmechanik, IJd. xxxiii. Heft 3 u. 4). 



65* 



