This set of two graphs aims to analyze polyrhythmic patterns of correlations between general history and press advertising. For that very purpose, we have first identified and listed in the related table a series of "conventional" events (i.e. recognized as "major" events in conventional chronologies of Shanghai, China and world history) (See the "Tables" section). For each event, we have then selected a relevant sample of the Chinese newspaper Shenbao and its British counterpart North China Daily (when available) - i.e. an edition published more or less shortly after the selected event. In order to measure the potential effect of each event on newspaper advertising, we have eventually established four material criteria:

• number of pages in the sample issue (grey line)
• number of advertisements (red line) (average advertising population)
• maximal area covered by a single advertisement (maximal advertising area), on a scale ranging from 1 (subdistrict, < 10% page) to 4 (empire, full-page advertisement) (blue line)
• average visuality (on a scale from 0 for purely textual ads, to 2 for visual advertisements, closer to the "model" copy) (green line).

Finally, the figures have been visualized on the following graph lines.

The first graph traces the lifeline of the Chinese newspaper Shenbao (1911-1949), while the second graph follows the trajectory of the British newspaper North China Daily News (1904-1951).

Four main observations emerged from these graphs:

1. The number of pages and ads appeared more directly affected than areas and visuality. 

2. The nature, degree, moment and duration of event impact may vary according to specific event. Events impacts are often unpredictable, if not unexpected:

• Some events seemed to have no impact at all: 1914, 1919, 1925, 1927 in the Shenbao; most of events before 1941 in the North China Daily News;
• Some events may have a very short impact (1912), while other had a very long impact (1937-1947);
• The event impact may be postponed or delayed: 1931, 1932, 1937 in the Shenbao;
• Some dramatic events may have contradictory effects: 1925, 1927, 1932-1935 economic depression are correlated to an unexpected newspaper and advertising expansion.

NB It is questionable whether or not contemporary events (for instance, 1925, 1927) can be credited for those changes in newspaper advertising (for instance, the spectacular ten-year expansion between 1925-1935).

What these graphs reveal are nothing more than correlations. Yet correlations are not evidence. As any visualization, these tables/graphs may only unveil invisible patterns, open new research questions, and possibly suggest alternative paths of interpretation.  

By juxtaposing the Shenbao and North China Daily News (NCDN) graphs, two further observations may be made:

3. The chaotic lifeline of the Shenbao suggests that the Chinese newspaper might be more sensitive to historical conjoncture than the British newspaper. In effect, the NCDN enjoyed a remarkable stability until the Second World War (more precisely after the US entered the war in Dec 1941). Then the NCDN so deeply suffered from the war and the restitution of foreign settlements that it never recovered, even though he survived the Chinese newspaper (which ended on May 27, 1949) until March 31, 1951. 

4. Despite its higher sensitivity to historical events, the Shenbao eventually managed to overcome recurrent troubles, recovering more or less rapidly after each crisis. That capacity to recover after each crisis may explain the oscillating appearance of the graph line.

NB The more cahotic line of the Shenbao may also reflects a distorsion in the method of sampling. As many editions of the North China Daily News are missing, it has not be possible to examine as many samples as we did for the Chinese newspaper Shenbao.