increased. It was for this reason that the current pole 

 and line were used in surface observations, as this 

 method is practically free from the above-mentioned 

 error. It may be possible to measure low -velocity sub- 

 surface currents accurately by placing the present type 

 of meter between a specially equipped anchor and 

 float, and allowing enough slack in the line to the boat 

 to prevent vertical movement of the meter. Time and 

 equipment did not permit such a procedure during this 

 survey. Furthermore, an accurate knowledge of these 

 low and variable velocities would be of value only in 

 connection with an extensive current survey. 



Subsurface current directions were measured with 

 a direction indicator, which consists of a carefully 

 balanced wheel, some 30 inches in diameter, suspend- 

 ed horizontally above the water and free to revolve 

 easily on ball bearings, and a 7-foot vane which is 

 supported horizontally and lowered to the required 

 depth by two fine wires which are attached to opposite 

 sides of the rim of the wheel (fig. 3). The vane, head- 

 ing into the current, moves the wheel correspondingly 

 by means of the wires. The direction is read on a sta- 

 tionary graduated circle attached to the axle of the 

 wheel. An ordinary bicycle wheel makes an excel- 

 lent indicator wheel. A lead -filled 1/2 -inch pipe, 

 split and flattened on one end to hold a 15 -inch piece 

 of sheet iron, serves as the vane. The wire is made 

 up in carefully measured pairs of the various lengths 

 necessary for the desired depths. A snap -hook is fitted 

 to one end of the wire and a small ring to the other 

 end. Two stiff wire rings are fitted through vertically 

 drilled holes in the vane pipe, thus enabling the wires, 

 which are snapped into these rings, to support the vane 

 properly. As the wheel is usually some distance above 

 the water, two light ropes, just long enough to span 

 this distance, are fitted to opposite sides of the wheel 

 rim and snap -hooks are bent on their lower ends. After 

 the vane is lowered to the required depth, the upfier 

 (ring) ends of the wires are attached to the above 

 snap -hooks at the surface of the water. 



In operation of the indicator, care must be taken 

 to lower the two ends of the vane simultaneously in 

 order to prevent the wires from fouling each other. 

 Also, it may be necessary to adjust the ropes, leading 

 from the wheel to the wires, slightly in order that the 

 wheel may remain perfectly horizontal. The gear is 

 supported on a 2- by 4 -inch outrigger to clear the 

 ship's hull. The indicator gives satisfactory results 

 when the current is greater than 0. 1 m. p. h. 



The sequence of work upon arrival at a station 

 was as foUows: First, a sounding was taken to check 

 the location. Second, top and bottom water sam- 

 ples for the chemist, top and bottom water samples 

 for plankton study, bottom (mud) sample, air and 

 surface water temperatures, and transparency were 

 obtained. When chemist's samples had been taken 

 the serial subsurface temperature measurements 

 were begun. Upon completion of the above, the 

 boat got under way at a slow speed and the 5 -minute 

 surface and bottom meter -net and the surface foot- 

 net hauls were made. If the station was to be oc- 

 cupied for currents, the boat was anchored upon ar- 

 rival and the net hauls made after weighing anchor. 



The time spent at a station varied from 20 to 

 60 minutes, depending upon the depth and whether 

 or not a thermocline was encountered, which neces- 

 sitated several more temperature observations. When 

 the boat was anchored for current determinations, 

 the time of occupation was from 1 to 2 hours. 



METEOROLOGICAL CONDITIONS 



Weather conditions, as shown by Weather Bureau 

 records, were about normal for the period of the sur- 

 vey except that the wind velocity was 9 percent above 

 the average. In table 4, which gives comparative 

 meteorological data, the figures are the means of the 

 Cleveland and Buffalo readings. The water temper- 

 atures, with the exception of those for May, are the 

 means of 37 selected stations fairly evenly distributed 

 over the lake, and offer an opportunity to study the 

 temperatures of the lake as a whole in relation to air 

 temperatures. It should be kept in mind, however, 

 that the water temperatures were taken during day- 

 light hours only, while the air temperatures are from 

 continuous day and night readings. 



There were two general storms during the period 

 of the survey when the wind velocity exceeded 54 

 m. p. h. One occurred on May 16 and the other on 

 September 10. On 11 days, the velocity exceeded 

 40 m. p. h. All of these blows were ftom the west or 

 southwest. The most stormy week of the survey was 

 from September 10-17, when the average for the Buf- 

 falo daily maxima was 46 m. p. h. The physical ef- 

 fects of these storms on the waters of Lake Erie are 

 taken up in detail in subsequent paragraphs. 



The prevaUing winds are westerly, and play an 



12 



