74 



FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 



By comparison, and in confirmation of the 

 results shown in experiment 2, while the mean of 

 the set held 61 days at subzero temperatures is 

 not significantly different from that of the initial 

 set, 4 of the 7 sets show differences with the t test 

 (table 4). The analysis of variance also shows a 

 possible significant difference between sets. 



From these results it became increasingly 

 evident that the time required to process the 

 samples was introducing an uncontrolled variable, 

 with changes in the ratio of inorganic to organic 

 phosphorus being more rapid than had been 

 suspected. The following experiment was designed 

 to control this. 



Table 4. — Changes in inorganic-phosphorus concentration 

 of sea water with added phosphorus, occurring at subzero 

 temperatures 



A. TALUKS IN Jl-ORAM ATOMS OF POl-P PER L1TEB 



B. COMPARISON OF MEAN OF INITIAL SET WITH MEANS OF 

 SUBSEQUENT SETS 



Time interval 



1 day.. 



2 days. 



3 days. 

 10 days 

 13 days 

 18 days 

 61 days 



Difference 

 between 

 means 



-0.19 

 -.29 

 -.19 

 -.20 

 -.05 

 -.25 

 -.05 



Value of t 



"2.914 



"4. 057 



•2. 225 



.016 



.828 



"4. 762 



.710 



Degrees of 

 freedom 



26 

 27 

 27 

 27 

 27 

 26 



F ='2.220 between sets 



Experiment 4 



Experiment 4 was designed to appraise the 

 effects of the delay inherent in drawing a large 

 number of samples on the concentration of 

 inorganic phosphorus. In this experiment, six 

 sets of samples were drawn as rapidly as possible 



directly from Offat's Bayou. Filtering was 

 eliminated to expedite the operation. In the 

 initial set, the molybdate reagent was added to 

 stabilize the samples, and the set was analyzed 

 immediately on return to the laboratory. Sub- 

 sequent samples, held at room temperature, were 

 analyzed at the intervals shown in table 5. 



The increases of 0.2 ix-gram atoms from 1.38 

 during tbe first 30 minutes and of 0.36 /x-gram 

 atoms within the first 90 minutes, and the sub- 

 sequent rise to 1.80 and decline to 1.68 p-gram 

 atoms, indicate how rapidly changes can take 

 place at room temperature and explain, in part 

 at least, the variations which developed in 

 experiments 2 and 3. It is important to note 

 that, while each set of samples was significantly 

 different from the initial, as shown by the t test, 

 and that the variations between sets in the 

 analysis of variance was large in relation to the 

 variance within sets, the standard deviations and 

 the variance within sets were nominal. The 

 importance of adequate stabilization or immediate 

 analysis in determining the inorganic-phosphate 

 ratio is here clearly demonstrated. 



Table 5. — Changes in inorganic-phosphorus concentration 

 of sea water occurring at room temperature 



A. VALUES IN ji-ORAM ATOMS OF POcP PER LITER 



B. COMPARISON OF MEAN OF INITIAL SET WITH MEANS OF 

 SUBSEQUENT SETS 



Time interval 



30 min. 

 BOmin. 



3 hrs... 



4 hrs... 

 6 hrs... 



Difference 

 between 

 means 



0.22 

 .36 

 .42 

 .40 

 .30 



Value of t 



"4. 422 

 ••8. 774 

 ••7.209 

 ••11.239 

 "7. 635 



Degrees of 

 freedom 



12 

 14 

 15 

 16 

 16 



