A possible cause is from roasting of pyrite ores which was frequently used as a source of sulphur for the sulphite process. The pyrite ores are commonly associated with significant amounts of arsenic. The arsenic may be release via exhaust gases or end up in the slag. This is a common environmental problem observed e.g. in Sweden.

Another possible cause is e.g. the production of wallpapers, which in the good old days sometimes were coated (at the back) with arsenic in order to withstand bugs and moulds.

A long story made short as:

[A]The grass/ trees/ cotton( to some extent)/ or any vegetation used in the production paper grow in arsenic-contaminated ground water and thus would enter some arsenic in them.And to add to it - we use Wood preservatives[ copper chromated arsenate (CCA) containing As2O3 and desiccants used in mechanical cotton harvesting.

[B] When such trees/ vegetation are used in the production of paper where chemical wood pulping involves the extraction of cellulose from wood by dissolving the lignin that binds the cellulose fibers. Kraft pulping is the major form of chemical wood pulping in USA, accounting for approximately 85% of pulp production.

[C]Four processes associated with the pulp and paper industry have been identified as potential sources of arsenic:

(I,ii)Chemical recovery furnaces and smelt dissolving tanks:Particulate emissions from the kraft recovery furnaces consist primarily of Na2 SO4 and Na2 CO3 , with some NaCl. Particulate emissions also contain arsenic, but only in minute quantities because arsenic is found as a contaminant in process chemicals and in trace amounts in wood.

(iii) Lime kilns :Emissions of arsenic from lime kilns are likely due to the arsenic content of the lime mud .

(iv)Power boilers: In addition, the pulp mill thermal chemical recovery processes associated with potential arsenic emissions,arsenic emissions from wood waste and fossil fuel-fired industrial power boilers are other sources of arsenic.

[D]The arsenic thus received remains dissolved in the liquid pulp as As2O3/As2O5/H3AsO3/H3AsO4 up to  a certain limit.

[E] Finally,it may well be noted that:

Particulate control and, therefore, arsenic control on recovery furnaces is always taken care of by the Chemists under the name “Arsenic emission factor” .

REFERENCE:

Fill/Click

presence of arsenic in paper industry

Reach/Click

LOCATING AND ESTIMATING

AIR EMISSIONS FROM SOURCES OF

ARSENIC AND ARSENIC COMPOUNDS Look in:

EMISSIONS OF ARSENIC AND ARSENIC COMPOUNDS FROM THE PULP

AND PAPER INDUSTRY 6.0-6.14.

It is possible that the wallpaper can be covered with arsenic for the purpose of giving durability

I would assume that when you state "guideline for the waste site investigation of former paper fabrication sites", you mean that you are working on writing guidance for the analysis of soil, sediment, surface water, and groundwater samples collected at either current or former paper manufacturing sites.  With that additional detail, it is possible to provide a clearer answer than you have already received.

It should first be noted that as arsenic is a metal, it will occur naturally in all soil, sediment, groundwater, and surface water samples that are collected at any site. If you have good enough detection limits (e.g., 0.1 mg/kg for soils or 0.1 ug/L for groundwater), you should always detect arsenic in almost all of your samples. Therefore, the most important question to ask when investigating a site where you are analyzing environmental samples for arsenic is "what concentration is naturally occurring in this area?"  For soils and sediments, the naturally occurring concentration will generally be above risk-based concentrations (e.g., risk based concentration for arsenic in soils from USEPA is 0.7 mg/kg but naturally occuring concentrations are generally in the range of 1 to 20 mg/kg).  Therefore, you must establish a relevant background concentration.  

For groundwater, concentrations above 10 ug/L (i.e., the MCL in the USA) can either indicate a release of arsenic or 1) low dissolved oxygen in groundwater or 2) low pH (both of which essentially leach arsenic from soils).  Thus, for example, a release of petroleum to groundwater can also result in elevated arsenic concentrations in groundwater as the biodegradation of the petroleum will lead to the leaching of naturally occurring arsenic from soils.

High arsenic concentrations in surface waters can also be indicative of either a release of arsenic or geochemical conditions that favor the mobilization of arsenic.

If, however, you find that arsenic concentrations in soils/sediments are elevated over background or in groundwater/surface water are elevated over background and not due to geochemical conditions (i.e., low dissolved oxygen or pH), then a release is possible.  Prior to modern synthetic chemistry, arsenic had a great many uses, as noted by the other commenters.  Arsenic (as lead arsenate) was used as an herbicide/pesticide up until the advent of DDT,  in the 1800s for wallpaper glue, in anti-fouling paint on boats, and in embalming fluid.  If the factory made their own acid to bleach paper, they may well have generated that acid by roasting pyrite ores (as suggested by Lars Olof Höglund).  The remnants of that process will be a purple slag that has very high arsenic concentrations.  However, that process has not been used since the 1800s (I think).

If there was an incinerator on site, the ash, bagasse, and/or slag from the incinerator may have elevated arsenic concentrations (as noted by Manohar Sehgal).  Further, the smoke emitted from the incinerator may have deposited arsenic downwind.  However, i would be more worried about dioxins/furans in soils downwind from an incinerator without proper/modern emissions controls.

If however, you were asking about the source of the arsenic in the paper products produced by a paper factory, the answer is as Manohar Sehgal stated:  arsenic is naturally occurring and the plants that were used to make the paper contained trace amounts of arsenic.  Similarly, arsenic is found in other plants that humans eat (e.g., rice).