40 CLIMATIC CYCLES AND TREE-GROWTH 



instrument consists simply of a slow-moving carriage on which lies 

 the wood sample and a fast-moving drum upon which hangs the record- 

 ing tape. These are connected by gearing which normally permits the 

 surface of the drum to move 12 times as fast as the carriage. A pair of 

 gears may be removed and another pair substituted, giving different 

 ratios, so that the range of magnification is from about 4 to about 34 

 times. In this way a convenient size is entered on the recording tape 

 and the record becomes partially standardized. The motion of the 

 carriage and specimen is watched through a small stationary telescope 

 placed a few inches above and f ocussed upon the rings and the motion 

 of the drum is recorded on the tape by a pencil line drawn across it 

 against a fixed wire. 



Accuracy — On the whole, an inexperienced assistant can handle 

 this plotter better than any other form of measuring instrument. 

 The duplicate records side by side check each other nicely. It is still 

 subject to errors of identification, but a large quantity of dated speci- 

 mens have gone through this process with good success. It is doubtful 

 if the settings have been quite as accurate as in the auto-plot, but they 

 are still as good as the sharpness of the rings permits. 



Graph and table — A sheet of coordinate paper is marked with dates 

 and then each ordinate is entered simply as the sum of the lengths 

 of that year in the two adjacent longitudinal plots. This gives an 

 ordinary graph on which a standardizing fine may be drawn, as 

 described below. Suitable ring values for entering in a group table 

 are then read off directly from the graph. 



CLERICAL OPERATIONS 



STANDARDIZING 



Need of equalizing trees — The groups of trees used in this study 

 represent different regions. Therefore, the individuals of each group 

 were selected to represent a considerable area rather than a localized 

 spot. Hence the individuals differ in rate of growth. What we want 

 in an average of a group is the common character which has come from 

 climatic variation. In the tables in Volume I a simple average was 

 used, as that was the easiest process and commonly used in scientific 

 reports. But it is perfectly evident that a straight average does not 

 represent an average of the common character, because in ordinary 

 averaging the big rings in quick-growing trees dominate and variations 

 in the slow-growing trees are practically lost. Logarithmic averaging 

 has been considered; for example, multiplying the values from the 

 different trees together and extracting the root corresponding to the 

 number of trees used. But that is a long and expensive process, and 

 it renders serious the occasional microscopic or omitted ring in very 

 slow growing trees. The effect in such cases would be greatly over- 



