THE THALIACEA. 



95 



After this there was a break of five clays in the observations, due to deep-water work 

 and to our running into Brest for coal ; except on the 19th and 20th we caught few 

 specimens, and the figures, both of average specimens per haul per day and of average 

 and actual lengths per day, are orderless. That sporadic sexual reproduction had 

 occurred subsequently to the first " swarm " is, however, seen from the fact that 

 sjiecimens of 2 mm. and less were caught (for the last time) on the 19th. 



We may conclude from the above : that a swarm of blastozooids may be formed, as 

 well as a swarm of gonozooids ; that it seems to demand the previous occurrence of 

 a swarm of gonozooids ; that some further explanation of the production of the swarm 

 is required than mere optimal conditions of food and temperature, because the swarm is 

 so local, even when sporadic reproduction is going on in the surrounding area. 



I had hoped that these blastozooids were present in sufficient numbers and occurred 

 on sufficient occasions to enable a test to be made of their reaction to light by the 

 method used for Sagitta serralodcntcda (Part III. pp. 60-61, above). But the attempts 

 at this were invalidated by the presence of the " swarm," which sent up the values of the 

 depths (50 and 100 fathoms) at which it occurred in a manner quite disproportionate to 

 those obtained for the remainder of the cruise. For example, the average specimens * 

 per hour haul at 100 fathoms between 3.0 a.m. and 7.0 p.m. were 93"5 if the swarm-haul 

 21 /i was included, but 32 if it was omitted! I therefore separated the swarm-hauls 

 21eto 21 /, and although they by themselves seemed to indicate a movement of the 

 sviarm to the surface at night, yet the placing of the remaining hauls on day-and-night 

 tables gave no satisfactory result : it neither confirmed nor denied any vertical oscillation, 

 either because both specimens and observations were too few, or, more probably, 

 because no such vertical oscillation exists in this species. But the contrasting of hauls 

 in dry weather, irrespective of time of day, with those made during or after rain 



* In this section the actual numbers caught have been treated with the time-factor; nets 18 and 180 omitted; 

 and averages struck over all hauls, not merely those in which the blastozooid occuncd. 



