The paper is well-written and the dataset will potentially be of interest to at least three groups of readers: climate research, ozone research, and science historiography.

For climate research, the dataset presents the interesting aspect of adding several decades of station data that predate what is taken as the acceptable observational record for surface ozone. The authors review the previously-published datasets and present appropriately the originality of the new dataset. The authors warn that exploitation will not be straightforward, for reasons cited in the paper, chiefly a strong dependence of the semi-quantitative ozone obtained by ozonoscope on humidity and other factors. The papers having reviewed the data quality from these instruments are also duly cited. However, to assit data exploitation, it seems humidity data are also available along with the observations. Furthermore, as data from other neighboring observatories and stations are rescued -in a similar way as done here- the accuracy of reconstruction of the humidity information is due to improve with techniques such as climate reanalysis (e.g., Slivinski et al., 2019, https://doi.org/10.1002/qj.3598).

One point to clarify would be if all observed elements reported in the original sheets got transcribed? or if some observed elements were not digitized, which ones were left out?

For ozone research, the dataset presents, besides monthly data, also daily nightly and diurnal observations (when available). This availability within the data series may present an opportunity to apply to earlier times our understanding of ozone chemistry and surface ozone variability, which is still developing (e.g., Monks et al., 2021, https://doi.org/10.5194/acp-21-12909-2021).

Finally, for science historiography, this paper presents the interesting point of illustrating how scientists today recognize, as a community, when observation techniques do evolve, and thus for some applications it may be preferable to leave aside some of the earlier records (e.g., Tarasick et al., 2019), but even then, how efforts continue to try and extract value from past observations (this paper).

Table 3: the image would need to be rotated (years are shown upside-down).

This paper presents a new dataset of semi-quantitative ozone concentrations in several locations of Iberia and western Africa for early periods.  Despite ozone measurements make with test-paper has been regarded as “not recommended for quantitative use” because of the sensitivity of test-papers to relative humidity, I consider that the data presented are very interesting and they deserve publication. Besides, the authors present a complete discussion about the characteristics of the data and its limitations, so future users are aware of the problems.

The paper is clear, concise and well written. The presented data are accessible and can be useful for future users, so I recommend publication. I only have some minor comments, mostly formal, that in my opinion could help to improve the clarity of the paper and the interpretation of the data.

Line 74. Consider adding a reference to Figure 1 here.

Lines 109-110. Here it is established that the ozone concentration was originally quantified in a scale from 1 to 11and then (line 118), it is said that this scale was later upgraded to one from 0 to 21. It is hard to know what system is used in the files and this is relevant for future users. Maybe labeling “[O3] Schönbein scale” or “[O3] Berigny scale” instead of “[O3] arbitrary units” in the data files would clarify the issue. In addition, in the paper it is said that the Schönbein scale goes from 1 to 11 (line 109) and the  Berigny scale goes from 0 to 21 (line 118). However in the files (at least in the ones I tested) it is said “This range of tonalities (function of the Ozone concentration) was 1 to 10 in the Schönbein scale and 1 to 21 in the Berigny scale”. Please clarify the discrepancies.

Paragraph from line 145 to 158. Consider adding a reference to Figure 2 for clarity.

Line 240. Do the authors mean tables 4 and 5?

Table 3. Please add grid lines for clarity.