47 



maximum occurred early, viz., in August and the spring one in 

 April. Although the general shape of the curve corresponds to 

 Dr. Lebour's, Ave find that the actual numbers of diatoms vary 

 very considerably. At Plymouth, the number of diatoms in 50 

 cc. of water at the autumn maximum was below 900, ours being 

 slightly in advance of this (1,100 in 50 cc), but while at the spring 

 maximum only 2,000 diatoms per 50 cc. were obtained at Plymouth, 

 our average was 8,000 per 50 cc. The number may, of course, 

 represent the numbers to be obtained in the sea at this time, but 

 it is possible the number in the tanks is higher than it ought to 

 be at this time as a result of resting spores formed in autumn 

 settling in the mud on the tank bottom and giving rise to numbers 

 of new diatoms in the spring. No attempt has been made to 

 prove this, but it seems reasonable in the case of Nitzschia clos- 

 terium at least, as quantities of small specimens were obtained 

 in April and May. Also, it must be remembered that the tanks 

 are not entirely emptied each day, so that a certain number of 

 diatoms will always be present from previous pumpings, together 

 with those pumped in each da}^ 



Another disadvantage of using the tank water fc/r this 

 inv^estigation is that large diatoms like Biddulphia spj). and those 

 occurring in tangled masses, e.g., Chaetoceras spp. evidently fail 

 to get into the tanks in anj^thing like representative numbers, as 

 these two genera occurred at rare intervals, and then only in 

 small numbers. This is evidently the case with other genera 

 also, for though the actual number of diatoms present in the 

 samples is high, the number of genera and species recorded is 

 much lower than those in the Plymouth samples or in Cleve's 

 work.* 



A list of the species of diatoms which occurred in the samples 

 is included in the present paper (Table I.) with the months 

 in which they are present, and a note as to whether they appear 

 as solitary individuals ( x ), are rare (r), in fair numbers (f ), common 

 (c), or very common (cc), and in addition to this the monthly 

 averages are given for the commonest and most frequently 

 appearing genera (Table II.), together with a number of others 

 which occur at odd times in fair numbers, or which are present 

 in greater or less quantity throughout the year. 



* p. T. Cleve's " Plankton Researches," Kongl. Svenska Vetenskaps-Akademiens Hand- 

 linger, Bander 36, No, 8. 



