That was for the U.S. only.
World-wide, we don't have anything approaching the measurements neccessary for a global calculation before WWII. See Climate Audit's "Where's Waldo" series of threads for examples of where the warming signal ISN'T - and where we don't have credible measuring stations.
Which is why it is absolute absurdity to be calling this a "trend" going back 100 years.
I challenge the concept of a "global temperature" measurement, let alone a "global temperature measurement trend" in the sense that is being used to push the AGW agenda.
Because the quantity and quality of data is not evenly distributed across the globe, you CANNOT say you have measured the "global temperature" and you certainly cannot string such unreliable data together to make a "global temperature trend" that means anything other than a curiosity. There are areas of the world overrepresented and underrepresented in the calculated average temperature. There are areas and time periods where "proxy data" is used from places like the USA to represent areas that did not have data. This means the place where the proxy data is used from is over represented, throwing the average calculation off of what actually existed. It is all meaningless. I'll believe a real trend taken over a century by an equally spaced global grid of well maintained modern sensors or a whole earth IR scan.
When the amount of temperature change that is being breathlessly trended is on the order of a degree or so, the magnitude of change so far is in the noise level and is highly subject to reading bias and historical changes of such biases. The world has gone insane.