You know, it is damned interesting to look at the temperature graphs today, and those of not so long ago. Here, have a look at this one produced by GISS in 1999
Note in your graph, in 1934, the temperature anomaly was .1 while in the graph produced by the same outfit in 1999 says that in 1934 the temperature anomaly was damned near 1.5...then in 1998 your graph says that the anomaly was about .2 while in 1999 the same outfit said that the temperature anomaly was about .75. Do you see anything wrong with the discontinuity of these two graphs produced by the same group using the same historical record?
Just wondering....