Why is blue darker than yellow in an analog black and white photograph? Blue is perceived darker than yellow by the human eye, because of biological principles within the eye. I can understand that therefore, when making a picture black&white in software like Adobe Photoshop, the software takes this perceptual information into account. 
However, in very old analog photographs, blue also appears darker than yellow. 
What is the reason for this? Is this just a lucky coincidence of the chemistry of the photographic film? Or is blue in some way really darker than yellow?
 A: When taking a black and white photograph, it is not uncommon to use a yellow filter to make clouds "pop" more in an outdoor picture (just as you might use a red filter to reduce the appearance of skin blemishes in a portrait photo).
Aside from that, the sensitivity of film varies - from the "panchromatic" film that tries to be "color neutral", to all kinds of other formulations that (de)emphasize some colors over others. This is described in some detail in this article which in turn borrows a lot from Ansel Adams' book "The Negative". Some of the facts from that article (do read the original):


*

*The first B&W film was sensitive in the near-UV through green: this is why such film could be developed by (dim) red light (ditto for photographic paper - but since that was exposed with "white" light through a B&W negative, the spectral sensitivity mattered much less)

*"Panchromatic" film arrived later - in this, the sensitivity was extended into the red range. However:



Early Panchromatic films were, however, still much too sensitive to blue light (what resulted e.g. in pictures with too bright sky and clouds invisible against white background), therefore required a yellow filter for correct representation of blue color brightness.

Over time, three types of panchromatic film were developed; they were thoughtfully named type A, B and C. Type A has increased blue sensitivity; type B has "response similar to the human eye"; type C which had extended sensitivity into the red, thus giving greater "effective speed" in artificial (tungsten) illumination.
The article goes on to mention more recent developments, including

Kodak T-Max films which have reduced blue sensitivity, thus the effect of tilting the curve toward red, but not an increase in red response or extension into infrared, this enables the response of the film to be closer to the response of the human eye, blues may be recorded as slightly darker tones.

In other words - whether you are getting "more" or "less" sensitivity depends on the exact film (and filters) you are using - there is no single rule that applies to all of black and white photography.
A: This will have some basic answers to your questions(which you have also sort of answered yourself; How is perceiving blue to be darker than yellow in Photoshop ´because of biological principles within the eye´ any different from perceiving blue to be darker than yellow in an old photo? Unless there is a way to ´see´ colors without using our eyes (or using existing preconceptions of what those colors actually look like) then blue will always appear darker than yellow when value and chroma are the same.
The ´biological principles,´ as you call them, are fairly complicated, with several different factors at play. One of these is called Color Constancy. This is basically how the brain perceives color under different lighting conditions. For an example of this, look at this photo, namely squares A and B.

Interestingly enough, A and B are the exact same shade. The context is enough for the brain to ´see the color differently.´ 
How this relates to different hues´ variance in perceived darkness is that the cones which feed the brain information which is contextually interpreted are not equally sensitive to all wavelengths of light. This will be further addressed at the end of this
The wavelength of peak spectral sensitivity of the eye in daylight is 555 nm, which is a green-yellow hue. At night/in dark conditions, that peak sensitivity shifts to 507 nm. You may notice that the environment looks more bluish at night, and this is in part due to the shift in sensitivity. There are images of this spectral sensitivity curve on the web, but I cannot load images at the moment. 
There are three different types of photoreceptor cones, often referred to as L, M, and S. They each have a unique peak sensitivity, with L(for long) having its peak at a wavelength of 560 nm. The M(for medium) cones have a peak spectral sensitivity at 530 nm, and S(for short) cones have peak sensitivity at 420 nm. 
These 3 types of cones are responsible for color vision, with wavelengths of visible light stimulating them all to varying degrees. Those proportional differences are what allow us to differentiate between colors under normal conditions. 
In a typical well lit environment, those cones that are most sensitive to the mid-spectrum wavelengths of light(M cones) are stimulated the most(since the light is primarily in that range of the spectrum,) which is where color constancy comes into play: Daylight brightness is a normative state for the environment, and all colors are perceived in that context. Since it is a normative state however, there is nothing to compare it to for ´actual normal,´ and everything else is only compared to that normative state. Color constancy has shown us that the same color can be seen differently with different color surroundings, and so in daylight those wavelengths furthest from daylight brightness will appear less bright(and thus more dark) due to the cones not being stimulated as much by the relative brightness.
