1 66 



SCIENCE. 



[Vol. XI. No. 270 



Fronteras, or San Bernardino, would have been demolished. These 

 houses stand in the centre of the section mentioned as having been 

 covered with water several inches deep, and which has sunk over 

 four feet. 



" Tombstone being the nearest place where a number of time- 

 observations could be compared, it becomes requisite to know with 

 as great exactitude as possible the instant the shock arrived. At 

 the onset of the temblor I had just noted the time, and my watch 

 was partially in my hand in the act of returning it to my pocket. 

 When I looked, it was 2.48 mining time, which was that day thirty- 

 five minutes faster than standard or railroad time. When, recog- 

 nizing the nature of the phenomena occurring, I again looked at 

 my watch, just one and three-quarters minutes had elapsed. This 

 was after the tremors had passed. The noise, resembling artillery- 

 firing more than any thing else, continued for some seconds, dying 

 slowly away in the north, to return in a few seconds from the south. 

 By careful comparison and consideration of at least thirty different 

 statements as to the Tombstone time, I am compelled to put the 

 onset at 2.13, with a possible and probable error of ten or fifteen 

 seconds. From Guaymas to Benson I have made personal inqui- 

 ries, and think there is room for more error than in the Tombstone 

 time. The difference in Sonora time and standard was that day 

 thirty-four minutes. 



"As concisely as possible I have tried to convey to you the facts 

 as seen by myself, and ventured some crude notions based upon 

 my interpretation of the observed phenomena. My idea of the ex- 

 planation of the opinion that the shocks came from the west is not 

 altogether theory. I have had the good or bad fortune to have 

 witnessed and felt a large number of mild and severe temblors the 

 past year ; and if the judgment of a non-expert, based upon per- 

 sonal experience, be worth consideration, then the assumption is 

 not valueless, and is worthy of your investigation." 



Mr. Goodfellow's report is accompanied by maps and photo- 

 graphs, references to which are omitted here, as have also been 

 some few brief passages not essential to a comprehension of the 

 facts. 



Distortion in Plane-Table Sheets. 



Mr. H. G. Ogden presented to the Philosophical Society, March 

 17, some observations on the effects produced in a plane-table 

 sheet by its absorption of moisture. It is well known that wetting 

 a sheet of paper causes it to swell or expand, and that, when the 

 sheet has dried again, it has returned to about its original dimen- 

 sions. Mechanical draughtsmen have realized the errors liable to 

 occur in their work from this cause, and have effected a partial 

 remedy by shrinking the paper upon a board, and securing the 

 edges to prevent further contraction, and then to provide against 

 expansion by working in a reasonably dry air. Surveyors using 

 the plane table resort to the same devices when executing work of 

 great refinement, especially if they have not the check afforded by 

 the positions of a triangulation previously plotted on the sheet. 

 But even with a well-conditioned triangulation, unless the paper is 

 glued to the board, the contraction is frequently the cause of great 

 annoyance, and sometimes of error. 



When a sheet of paper expands from moisture, the percentage 

 of increase in length is less in the direction of the grain of the 

 paper than at right angles to that direction, or across the grain, and 

 the difference between these percentages is practically the distor- 

 tion. If the percentage of increase should be the same in both 

 directions, there would be only a change of scale : all distances 

 would be proportionally increased, and the points would bear the 

 same relations to each other as before the expansion. 



While it is true that the primary cause of the distortion is the ab- 

 sorption of moisture, the resultant effect is usually a permanent 

 contraction, subject only to slight changes, except under the condi- 

 tion of excessive moisture. It is while this point of permanent con- 

 traction is being reached that the greatest annoyance is experienced 

 and the most serious errors are likely to occur. Mr. Ogden then 

 described a series of experiments made at the Coast Survey office 

 some years ago, to ascertain the changes in length that took place 

 in the hand-made antiquarian paper backed with muslin. Inspec- 

 tion of these observations, when plotted in the form of a curve, 

 shows that there was a decided tendency to assume a state of per- 



manent contraction. The readings at this point for each cross- 

 section are 35 and 26 d. c. m., or at the rate of a little over 

 .5 d. c. m. per inch of paper in one case, and .4 d. c. m. in the 

 other. The difference between these rates is the average distor- 

 tion in this case, — a quantity that is quite appreciable in a foot 

 of paper, and very plainly so in two feet. In field-work, Mr. Ogden 

 said that he had frequently found the distortion double this amount, 

 and in one instance he remembered it nearly three times as large. 

 He had also worked a whole season without any appreciable devel- 

 opment of distortion. 



In charts or maps printed from copperplates, or by any process 

 that required wetting the paper, this distortion is largely developed, 

 usually averaging as great as one per cent, and, if inferior paper 

 has been used, as much as two or three per cent. The fact that 

 this distortion exists is frequently not realized even by the most ex- 

 pert navigators, and some even magnify the errors by laving off 

 courses with a protractor, discarding the compass-cards printed 

 with the chart. These compass-cards are affected by the contrac- 

 tion exactly as the sheet is : directions ascertained by them are per- 

 fectly good, but the graphic angle between any two directions is 

 erroneous, except in the case where the directions are on the lines 

 of contraction. 



Mr. Ogden then referred to a diagram, a copy of which is given 

 herewith. A, D, C, etc., were plotted in the true relations. After 



the sheet has contracted, a, b, c, etc., represent the relations those 

 points have assumed. The paper contracts at a uniform rate in 

 each direction. 



The table is supposed to be at X, the exact centre of the figure, 

 and it is required to determine the position by the distorted points 

 a, b, c, etc. By reversing the telescope, we immediately ascertain 

 that we are directly on the line HD. It will also show that we are 

 on the lines AE, CG, and BF. But the distortion is not apparent 

 until the telescope is pointed at the signals, and the lines are 

 drawn on the sheet. Then if we orient by the line HD, we shall 

 produce the figure of the diagram, giving five determinations, i, 2, 

 3, 4, and .'Y,each made with four well-conditioned points. Any one 

 of these positions would be considered satisfactory if we had not 

 the other points to show that something was wrong. To orient by 

 the lines BF will produce the same result. But if we take the diag- 

 onal AE, we shall have two positions at 5 and 7, formed by the 

 intersection of the diagonal points, with the lines from the other 

 points running wild. Using the diagonal CG would give two 

 points at 6 and 8, with the lines at the other points running wild, as 

 before. 



There is no question, that, out of the nine positions developed by 

 these settings, that at X is the only true compromise. When the 

 sheet is distorted, all'positions are compromises ; and A' is the true 

 compromise in this case, for it is on the lines GC, AE, etc. : a being 

 below, and e above, the line connecting A and E, by equal quanti- 

 ties, a line drawn between the distorted points a and e must pass 

 through the middle point X. The positions 5, 6, 7, and 8 cannot 

 be true, because the lines forming them will not pass through the 



