to attempt to plot the run of the vessel upon it with any greater degree of nicety than has been used in ascertaining the run by the means that have been used; that is to say, by means of instruments, etc: The inaccuracy of the variation is one source of this error. OBSERVATIONS FOR VARIATION, Most of the observations for variation, especially those of the open lakes, were taken a good many years ago, when: in- struments of precision were unknown, to say nothing of the other errors of ob- servation. Again, the annual change in the variation is a quantity not to be re- "TAE Marine REVIEW from ranges, recollect that if the shore bearing is to right of the compass bear- ing, the deviation is easterly, and if to the left, westerly. ON SHORT RANGES. It is a very difficult matter to place a ship's head exactly on a short range and _ keep it there for any length of time. This is owing to the short distance be- tween the ranges, making it possible for the ship to swing two or three degrees without. being detected. Another thing, no two officers will agree as to when the ship's head is on the range. This is often due to the fact that they do not view it Showing the effects of parallax of the compass when a ship is swung in a large circle and short distance from object as compared with a small circle and a great distance from object. O O objects on shore; A B D E circle the ship is swung in. The lines drawn from them to O O show the angle of parallax due to the closeness or farness of the objects from the ship, and also a comparison of the size of the,circles the ship describes while observing bearings of: OO. lied .upon, simply because there is no regularity to the change, and the change found for one station would not agree with the change for a second station even though it be not far removed' from that of the first. With 'the rough instruments (rough compared with those of the pres- ent time) then in use, together with the faulty methods of observation, the inac- curacy of this problem must be apparent. It is safe to say that no observation is closer than the even degree. The annual change is determined by observing the variation after a lapse of years; the dif- ference divided by the number of years. that have elapsed, gives the annual change. Probably, the variation had only changed a half degree during the inter- val, an amount hard to detect with the instruments then in use. This is the man- ner in which the variation is computed and kept corrected in certain places. The - variation thus found does not agree with the variation as found by observation, after a lapse of years. Considerable dif- ference is found in the observed variation and that of the computed variation. The government is and has been establishing more accurate data on this subject, both in establishing the correct amount and also in correcting the annual change. There is-still a great amount of naviga- ble territory on the lakes to be covered by such observations. This variation sub- ject is of inestimable value to the lake navigator, and he is just beginning to ap- preciate its service now that the Time Azimuth mode is coming into general use on the lakes. TO NAME THE DEVIATION. In naming the deviation ascertained from the same line of position. The wheelsman usually stands to one side or the other of the wheel, and in placing the' vessel's head on the range, he does not see it the same as he would were he to. stand directly over the center of the wheel. THE DEV.-FOR SHIP'S HEAD AND, NOT FOR THE BEARING, Be careful when correcting a bearing, not to commit the blunder of taking the deviation from the bearing point. It must, of course, be taken from the one which corresponds to the ,deviation of the ship's head at the time of observa- tion. This mistake has frequently been made. Deviation always refers to the ship's head. In other -words, where deviation is given it means for a certain course. If the deviation table or curve tells us that the deviation on NE is 8° Wly, it signifies that when the ship's head is NE there is that amount of deviation. CARE OF BEARING INSTRUMENT, See that the bearing instrument is in good condition, and before 'using, it should be examined to see that the at- tachment revolves freely and smoothly in every direetion. If it be an azimuth at- tachment, see that it does not throw the compass bowl out of level. The levelling bubble on the bar of the attachment will _ show whether or not the instrument is plumb. GENERAL OBSERVATIONS, In the foregoing processes for obtain- ing the deviation of the compass it has been assumed that. every facility for . swinging the ship has been available, but as circumstances may render inconveni- 29 ent the making of observations in detail from point to point of the compass, there will be 'found below the description of a graphic method whereby from the devia- tions determined, as the ship swings round at anchor, or otherwise, on any. di- rection of the ship's head a curve of de- viations of the thirty-two points may be constructed afid the deviation determined for each point of the compass. This method has the further advantage that any number of observations made in the complete circuit of the ship may be made use of and the accuracy of the curve and the table proportionately increased. GRAPHIC METHOD OF FORMING TABLE OF DEVIATION, On a sheet of accurately ruled paper draw a vertical line down the center, representing the magnetic or correct mag- netic, North and South. The equidis- tant horizontal lines may be taken as points of the compass on the vertical scale, and the degrees of deviation on any convenient scale measured from them _tight and left from the vertical line. Easterly deviations should be plotted on the right, westerly on the left. A flowing curve being drawn through each point plotted on the paper, the de- viations for every direction of the ship's head may be determined by measuring from the vertical line. : The symmetry or otherwise of this _ curve will indicate (in the case of ves- sels where the iron is regularly disposed round the compass) the correctness of the deviations obtained; also the amount of agreement between the flowing curve and the line traced by joining each point 'plotted will show if any individual obser- vations have been incorrect. -- Se Should any point be omitted in swing- ing the ship the deviations on them can be measured from the curve drawn as above. noe From the observations thus made, or by either of the processes which have been described, a table of the results should be forthwith constructed for gen- _ eral use and copied by every person on board who keeps.a reckoning. . SHIP'S HEAD BY COMPASS:OR CORRECT MAGNETIC. It must be remembered what has been said about the deviation corresponding to the ship's head by compass and the ship's head correct magnetic. In all this work, except where the vessel heads di- rectly on a range whose correct magnetic © bearing is known, the deviation corre- sponds to the ship's head as indicated by compass. In running the ranges of the St. Mary's and Detroit rivers, coming in and going out of harbor piers, the devia- tion can always be obtained for the ship's head correct magnetic instead of ship's head by compass. In Azimuth Work we