General

This paper makes a well-argued case that the record Irish (Eire) temperature should be the record for Boora in 1976.and not Kilkenny in 1887.  However, I have a significant suggestion that could make this finding either stronger or weaker, or just possibly, change the decision. This involves considering the evidence from the several Figures in the paper like Figure 4. Here 500mb heights are shown in addition to the very useful pressure at mean sea level (PMSL) maps. It would be considerably more useful to replace 500mb heights, difficult to interpret in the context of this paper and not made much of anyway, by a relatively low-level reanalysis temperature field that is unlikely to be corrupted by a bad surface air temperature observation. I come back to this in the specific comments below   

The other main comment is that aspects of the presentation should be improved, particularly many of the remaining diagrams.  Nevertheless, the text is generally well written.  I recommend this paper be accepted subject to significant/major revision.

Specific  Comments

1. Lines 30- 62 of Introduction and/or Table 2. Somewhere about here it would be very useful to add the latitude, longitude and height location of Kilkenny Castle. In Table 2 it would add useful information to add these details for all the other stations shown –much as done in Table 3.
2. Figure 1. It is not so easy to see on a printed copy where Kilkenny Castle is located on either map. Perhaps an arrow and superimposed text of suitable size saying “Kilkenny Castle” could be added.
3. Figure 2. The coloured dots would be easier to see if the green background was much lighter.
4. Figure 3. This diagram is not very clear on a printed copy, mainly because of the grey background. This can be fixed by making the graph lines thicker.  The key is odd: the heading should be “stations” not “variable” and Kilkenny has a coloured line in the key, even though there are no data for this date, except the single point highlighted in the text. On the left, the label should add “degrees” before “Celsius”. In the caption “air” should be added after “surface”.
5. Figure 4 and similar Figures. This is my most important comment. On the main Climate of the Twentieth Century Reanalysis v3 NOAA web site atmospheric temperature analyses are available in ensemble mean form for 8 times of the day for each day until 2015. I strongly recommend that 500hPa height be replaced by an appropriate relatively low-level temperature (850hPa temperature or lower, as long as not significantly affected by station biases). This would be done (probably) for 1500 hours for each sub-diagram shown in Figure 4 and the similar figures. Ensemble mean PMSL would need to be be plotted for the same time of day for consistency. This would add genuine new information about the relative likelihood of the hottest temperature extremes. This type of information could become an important part of the evidence for daily surface air temperature extremes more generally. As far as I can see, appropriate sub-daily maps can be easily plotted for such data using the on line facility at https://www.psl.noaa.gov/data/20thC_Rean/ . After 2015, similar data should be available from ERA5.
6. Figure 5. On a printed copy, the red circle is less clear on some sub-diagrams than it should be. These are good diagrams otherwise. It might be useful to add combined simultaneous differences for all stations in a sixth panel -currently blank.
7. Figure 7. Where relevant, I have the same comments about the graphs as for Figure 3.
8. Figure 10. Same comments about the red dots as for Figure 5.    

   Minor comments

1. At Line 241, Parker (1994) is missing from the Reference list.

This analysis reassess the 1887 high temperature record of Kilkenny, Ireland, using different datasets including the early-year records, current observations and reanalysis product, and concludes that this record and other early observations of extreme high temperature are not sufficiently reliable. The authors instead suggest the 1976 heat record at Boora as the highest national temperature measurement. The analysis is interesting, and the conclusion is also more or less convincing. This is not an easy work anyway. One of my concerns is the locality of the study, and thereby the narrowness nature of its significance if it is published in the CP. In addition, I find a few of other issues which demand further clarification.

1. The Kilkenny observational site is located in the centre of Kilkenny Town. Was it also located in the town centre in 1887? What was the extent and population of the town in the early time? If the town has not changed much in urban area and population, would the central location of the observation in 1887 be more productive to the higher temperature records compared to other observational stations in the nearby cities and towns of similar size?
2. The application of 20CR to show the synoptic situation seems not so persuasive, I feel. It indicates a favorable weather condition for the extreme hot and dry event, but when the authors said that the condition is not uncommon in the summer season over Ireland, they did not exhibit an example of the synoptic situation at present. Is there any similar or even more extreme high pressure system over the study region during the same length of periods in modern time?
3. The exercise to compare the JJA season differences of maximum temperature between Kilkenny Greenshill and the four stations is convincing procedure. However, the representativeness/proximity of Kilkenny Greenshill to Kilkenny Castle would be important. The authors did not well document this. It would be good to at least give a large-scale map showing the specific locations of the two sites. Besides, all the data used should be evaluated for their quality and homogeneousness.
4. Given Kilkenny Greenshill could well represent Kilkenny Castle, there are other issues that should be clarified. 1) Birr seems mostly close to Kilkenny Greenshill, with both being inland stations, but its modern data was not shown, why? 2) For the two sites (Roches Point and Armagh) which is located inland or more closely to Kilkenny, the differences of Tmax between 1887 Kilkenny Castle record and those of the two stations are within 99% of the modern data, which means that the differences are not impossible. If the reference sites are closer to Kilkenny, would the possibility for the 1887 Tmax differences to occur be higher?
5. Finally, I would suggest that all the figures be improved for their quality and clarity.