OCEAN TEMPERATUEBS. 169 



temperature curves at clififerent stations will be observed on even a very superficial comparison of 

 the charts. It will also be noticed that, while at the southern stations the temperature curves are 

 generally most regular during the summer months, the reverse is true of the extreme northern 



ones. 



« 



THE SUEPAOE ISOTHEEMS. 



The purpose in view in preparing the charts of surface isotherms (Noa. 26 to 31), has been to 

 present the temperature observations on which this report is based, in what appears to be the 

 most convenient form for use in connection with such fishery problems as are suggested by the 

 migrations of surface schooling fishes. If such species as the mackerel are controlled in their 

 movements toward the north by conditions of temperature that are constant for all latitudes, a 

 line drawn upon a chart to indicate their progress with reference to time must agree more or less 

 closely with some line of equal temperature projected from point to point along the same coast. 

 This supposition expresses in a general way the belief of many persons who have studied the 

 migrations of mackerel and other "economic fishes, but up to the present time sufficient data have 

 not been collated to render possible the practical application of the principle to those species that 

 live solely in salt water. 



The isothermal charts are seven in number; one for each of the five years from 1881 to 1883, 

 inclusive, one representiug the means of the same five years, and the final one illustrating the 

 relations of the air and surface isotherms. The annual charts are of most importance for fishery 

 purposes, as, in showing the changes of position Of the isothermal lines from year to year, they may 

 possibly serve to explain the causes of the irregularity in the appearance of certain species upon 

 different parts of the eastern coast in different years. A few words of explanation are necessary 

 respecting the construction and contents of these charts. 



Although, as elsewhere explained, the temperature results are not of equal value at all the 

 stations, the latter have all been included in the charts, for the reason that it was impossible to 

 determine satisfactorily, excepting in a few cases, which, should be excluded. The observations 

 at Cape Lookout, Body's Island and Absecon Inlet evidently do not apply to the open waters of 

 the coast, and the same is probably true to some extent with respect to a few of the other mainland 

 and island stations. The three stations specially referred to have not generally been considered 

 in discussing the isothermal charts. The data for the construction of the charts has been taken 

 from the original temperature records, and not from the reductions to ten-day means, although the 

 latter have been considered in deciding every date used in constructing the isotherms. In deter- 

 mining the dates for each isotherm no observations were considered unless the means of twenty 

 consecutive observations (ten days) equalled or exceeded the temperature of that isotherm, except 

 in a few instances elsewhere explained. The temperature of 40°, for example, might be reached at 

 any station either within the first ten-day period indicated upon the special chart of that station 

 (charts of ten-day means) as having a mean of more that 40°, or in the latter part of the previous 

 ten-day period; but the mean temperature for the ten days following and including the date of the 

 isotherm must not be under 40°. For the isotherms during the period of ftiUing temperature in 

 the last half of the year this order is reversed. 



In constructing the charts, the names of the stations have been arranged vertically, in geo- 

 graphical sequence, on the left hand side of the chart. The remainder of the chart is divided into 

 thirteen vertical spaces, each representing one month, that on the extreme right being for the 

 month of January of the year following that to which the chart relates. Each month is further 

 divided by the fainter lines into five equal parts, for convenience in reading the dates. The iso- 



