This figure in Jaworowski’s statement is reprinted from his article in The Science of the Total Environment 1992, and his related argument seems as odd as it did then. About the figure he says:
In Figure 2 encircled values show a biased selection of data used to demonstrate that in 19th century atmosphere the CO2 level was 292 ppmv.
Figure 2 has been widely reprinted:
Now let’s pause a moment to let Jaworowski’s description sink in. Look at that cloud of data points pre-1900, and compare those to the more modern values. Without knowing more about the data, it seems reasonable to suspect that atmospheric CO2 measurements before 1900 were highly variable, and may have been systematically high, compared to modern measurements. Jaworowski is trying to say that rejecting the outliers shows “a biased selection”! This seemed so preposterous that I looked up what he and his colleagues said in their 1992 paper. Yes, they actually do make that claim:
Slocum  pointed out that, from a set of twenty-six 19th century averages, Callendar rejected 16 that were higher than the global average of 292 ppm, and only two that were lower. On the other hand, from the 20th century data set Callendar rejected three averages that were lower than his global average of 317 ppm, and none that were higher. This shows a bias in the selection method.
In fact, Jaworowski et al. get little comfort from Slocum, who in his 1955 paper
All this does not refute Callendar’s thesis. The available data merely fail to confirm it. The positive evidence that the CO2 has increased is inconclusive, but seems strong enough to reward further study, and the time seems ripe for new research.
In the intervening 50 years, new research has been performed that nicely vindicates Callendar’s selection criteria, as Tim Lambert showed by superimposing data sets from Antarctic ice cores and atmospheric measurements taken at Mauna Loa. Both data sets independently confirm the trend observed by Callendar. They were available to Jaworowski when he prepared his statement, but he ignored them.
Jaworowski’s accusation of bias is surprising, just taking the evidence on its face. But let’s go the extra mile.
A reasonable person, trying to be fair, might wonder if the paper by Fonselius et al., whence the figure originated, had suggested some bias in data selection. OK, let’s have a look at Fonselius. Here’s the original figure:
Note that the sources of the selected data are identified; these are discussed in the paper. Now, did Fonselius et al. suggest that Callendar’s choice of data was biased, and that he should have taken an average of the whole cloud of data points from before 1900?
You can probably guess the answer to that one. No, they didn’t. They said:
This scatter … probably depends on the rather crude and incomplete techniques of analysis used during the first fifty years, possibly with systematic positive errors.
Fonselius et al. go on to discuss how CO2 measurements might be improved and collection extended, remarking that
In the diagram, the values used by Callendar and our mean values are encircled and we can see that our values fit in quite well.
Yet, Jaworowski has taken the figure from this paper, redrawn it, and used it to support his argument that Callendar used a “biased selection” in his choice of data. Now, that’s chutzpah!
 Slocum, G., Has the Amount of Carbon Dioxide in the Atmosphere Changed Significantly Since the Beginning of the Twentieth Century? Monthly Weather Review, December 1955, 225-231
 Fonselius, S.; Koroleff, F. and Wärme, Carbon Dioxide Variations in the Atmosphere, Tellus VIII (1956), 176-183